A method for identifying and assessing drought risk coupled with vegetation evapotranspiration dynamics

Abstract

The present application belongs to the technical field of meteorological disaster monitoring and risk assessment, and discloses a sudden drought identification and risk assessment method coupled with vegetation evapotranspiration dynamics. The method comprises the following steps: obtaining actual evapotranspiration data and potential evapotranspiration data and performing quality screening and mask processing; calculating evaporation stress ratio and performing standardization processing to obtain a standardized evapotranspiration stress index; calculating the difference of the standardized evapotranspiration stress index in adjacent preset time periods in the to-be-evaluated period and standardizing it to obtain a standardized evapotranspiration stress index change amount; scanning the standardized evapotranspiration stress index change amount, and identifying a candidate sudden drought event when it is lower than a percentile threshold of the initial stage of drought; and applying duration, end stress state and overall change rate constraints to determine an effective sudden drought event. The technical problems of the prior sudden drought identification method, such as lack of coupling with vegetation evapotranspiration dynamics, low identification accuracy, inability to comprehensively evaluate sudden drought ecological risk, and lack of post-disaster vegetation recovery force evaluation, are solved.

Description

Technical Field

[0001] This invention relates to the field of meteorological disaster monitoring and risk assessment technology, specifically to a method for identifying and assessing sudden droughts by coupling vegetation evapotranspiration dynamics. Background Technology

[0002] Flash drought, also known as sudden-onset drought, is a type of extreme drought event characterized by rapid onset, short development time, and high destructiveness. Its outbreak is typically accompanied by precipitation deficit, abnormal high temperatures, increased evapotranspiration demand, and rapid decline in soil moisture, significantly impacting agricultural production, forest ecosystem functions, and regional water security. Compared to traditional slow-onset drought, flash drought is more sudden and insidious, making it difficult for conventional monthly drought indices to capture its rapid formation process in a timely manner, thus posing significant challenges to subsequent risk warnings and ecological impact assessments. Existing research indicates that abnormal evapotranspiration, vegetation water stress, and soil moisture deficit are key indicators of flash drought occurrence and development. Methods such as the evapotranspiration stress index and standardized evapotranspiration stress ratio can comprehensively reflect the drought status of the soil-vegetation-atmosphere integrated system by analyzing abnormal changes between actual and potential evapotranspiration. However, in forest ecosystems, vegetation phenology, canopy structure, root water uptake capacity, and stomatal regulation processes alter the intensity of evapotranspiration response to atmospheric drought and soil moisture deficit, thereby affecting the identification time, development speed, and severity of flash drought, leading to biases in ecological risk assessments.

[0003] Currently, scholars both domestically and internationally have proposed various methods for identifying surge droughts based on soil moisture or the Standardized Precipitation Evapotranspiration Index (SPEI). However, an ideal technical solution that can fully couple vegetation evapotranspiration dynamics while simultaneously considering surge drought identification and comprehensive risk assessment is still lacking. Especially given the combined impacts of forest management and other anthropogenic disturbances on forest ecosystems, as well as climate change, accurately characterizing the role of vegetation evapotranspiration changes in the formation, evolution, and risk amplification of surge drought events remains an urgent problem to be solved.

[0004] The method for identifying flash droughts based on rapid soil moisture decline is a common technical approach in current flash drought monitoring research. Its process typically includes: acquiring root zone soil moisture data, standardizing time scales, converting percentile thresholds, determining the rapid decline phase, constraining drought persistence, and identifying event termination. Specifically, this type of method utilizes root zone soil moisture data from land surface hydrological or reanalysis data such as the Global Land Evaporation Amsterdam Model (GLEAM) and NASA's Global Land Data Assimilation System Version 2 (GLDAS-2). Daily-scale data are standardized to a preset time scale (every eight days), and multi-year historical soil moisture percentiles are calculated for each spatial unit and intra-annual period. In the event identification process, the core criterion is usually a rapid decline in soil moisture from a relatively normal state to a drought state. For example, it requires root zone soil moisture to decline from above the 40th percentile to below the 20th percentile at a certain rate, while maintaining a low soil moisture state for a certain period to eliminate misjudgments caused by short-term moisture fluctuations. When soil moisture rises back above the drought threshold, the flash drought event is considered to have ended. The advantage of this type of method lies in its relatively intuitive physical meaning, reflecting the process of rapid water deficit in the soil layer, and its applicability to agricultural drought and root zone water stress monitoring. However, this method still has significant shortcomings: on the one hand, its identification results are highly dependent on the soil moisture dataset, soil depth setting, percentile threshold selection, and soil moisture response characteristics to precipitation pulses, which can easily lead to differences in event frequency, duration, and intensity identification between different data sources and different climate zones; on the other hand, soil moisture essentially mainly characterizes the state of groundwater and is difficult to directly reflect ecophysiological processes such as vegetation canopy transpiration, stomatal regulation, phenological changes, and root water uptake strategies. Therefore, it may not be able to accurately characterize the true response of vegetation to water deficit under different vegetation cover types such as forests, grasslands, and farmland. In addition, existing soil moisture products generally suffer from problems such as coarse spatial resolution, large uncertainty in root zone water inversion, and insufficient applicability to complex terrain and heterogeneous vegetation areas, especially in capturing the response to sudden drought caused by local ecological changes such as forest gain / loss, vegetation degradation, and human disturbance. Therefore, relying solely on a rapid decrease in soil moisture to identify sudden droughts cannot adequately reflect the characterization of vegetation dynamics, surface energy exchange, and land-atmosphere coupling processes.

[0005] With the development of remote sensing observations, land surface models, and reanalysis data, actual evapotranspiration, potential evapotranspiration, vegetation indices, radiation flux, and meteorological driving data are increasingly being incorporated into surge drought identification and risk assessment studies. Among these, vegetation evapotranspiration dynamics are receiving increasing attention, closely related to its variable characteristics. This process is located at a critical interface of the soil-vegetation-atmosphere continuum, simultaneously reflecting the coupling and interaction of surface water supply, atmospheric evapotranspiration demand, and vegetation physiological regulation. Compared to soil moisture, which primarily describes changes in groundwater storage, evapotranspiration and its stress states more closely approximate the actual drought response of vegetation, comprehensively reflecting the combined effects of multiple factors such as high temperature, low humidity, enhanced radiation, wind speed changes, precipitation deficit, and insufficient soil moisture on vegetation water use processes.

[0006] Recent studies have proposed new indicators, such as the Standardized Evaporative Stress Ratio (SESR), which attempts to characterize whether land surface water supply can meet atmospheric evapotranspiration demand by utilizing the relative relationship between actual and potential evapotranspiration. This provides a more direct reflection of vegetation transpiration limitation and the degree of water stress in the ecosystem. Compared with the traditional soil moisture threshold method, introducing vegetation transpiration helps eliminate differences in vegetation types, enhances the comparability between regional and phenological stages of vegetation, and is more conducive to revealing the coupling relationship between "insufficient water supply—intensified atmospheric drought—limited vegetation transpiration" during sudden droughts. Especially in forest ecosystems, vegetation canopy structure, root depth, stomatal regulation, and phenological stages significantly alter the process of soil moisture deficit transforming into ecological stress. Relying solely on soil moisture often fails to reflect vegetation function impairment and changes in land-atmosphere coupling intensity in a timely and accurate manner. Therefore, although existing drought identification technologies based on soil moisture have certain basic and universality, they are still insufficient in characterizing vegetation dynamic response, identifying ecological risk amplification processes, and supporting high-precision risk assessment. There is an urgent need to further introduce key factors that can characterize vegetation evapotranspiration changes and land-atmosphere coupling processes to improve the ecological interpretability of drought identification and the reliability of risk assessment.

[0007] The technical solution described in patent number CN110334404A proposes a method for accurate identification of sudden drought at the watershed scale. This method constructs a watershed VIC (Vascular Indicator) land surface hydrological model, outputs variables such as soil moisture content, and identifies sudden drought events and classifies their risk levels accordingly. While this method can identify sudden drought within a specific watershed, its implementation is highly dependent on the land surface hydrological model, requiring steps such as model construction, parameter calibration, input data preparation, and result verification. This results in a high technical threshold, high model computational complexity, and high costs for widespread application, leading to insufficient generalizability. Furthermore, this approach primarily focuses on watersheds, resulting in a limited application scale and scope, making it difficult to directly apply to sudden drought identification across global or large-scale vegetation ecosystems. It also lacks characterization of the ecological impacts of sudden drought, such as crop yield reduction, forest degradation, and decreased vegetation productivity. Therefore, this existing technology suffers from high model dependence, difficulty in widespread application, limited research scope, and a lack of ability to identify vegetation ecological impacts.

[0008] The technical solution in patent number CN112395557A proposes a method for identifying sudden drought based on instantaneous development rate. This method focuses on using the development rate of drought events within a short period to determine sudden drought events, reflecting the rapid development characteristics of sudden drought. However, this method focuses on a relatively singular feature of sudden drought, easily misclassifying short-term water consumption during normal vegetation growth as sudden drought. For example, during the leaf expansion stage or rapid growth period, increased vegetation transpiration leads to a short-term decrease in soil moisture, increasing the instantaneous development rate. However, this process may be a normal phenological change and does not cause substantial drought stress to the vegetation. Identification based solely on the instantaneous development rate may lead to an overestimation of the frequency of sudden droughts, reducing the accuracy and ecological interpretability of the identification results. Furthermore, this method lacks further assessment of the degree of vegetation damage and ecological risk level after a sudden drought, making it difficult to directly support practical applications such as agricultural production, forest protection, and disaster early warning. Therefore, this existing technology suffers from problems such as a single identification indicator, potential overestimation of event frequency, insufficient consideration of vegetation dynamic response, and limited risk assessment capabilities. Summary of the Invention

[0009] The purpose of this invention is to provide a method for identifying and assessing the risk of sudden drought by coupling vegetation evapotranspiration dynamics, which solves the technical problems of existing sudden drought identification methods, such as lack of coupling of vegetation evapotranspiration dynamics, low identification accuracy, inability to comprehensively assess the ecological risk of sudden drought, and lack of assessment of post-disaster vegetation resilience.

[0010] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics includes the following steps: Step 1: Obtain actual evapotranspiration data and potential evapotranspiration data for multiple consecutive preset time periods within the target spatial area, and perform quality screening and masking processing on the actual evapotranspiration data and potential evapotranspiration data; Step 2: Based on the filtered actual evapotranspiration data and potential evapotranspiration data, calculate the evaporation stress ratio of each spatial unit within the target spatial region; Step 3: Obtain the mean and standard deviation of the evaporation stress ratio for the target spatial region over many years corresponding to a preset time period. Then, use the mean and standard deviation of the evaporation stress ratio to standardize the calculated evaporation stress ratio and obtain the standardized evaporation stress index. Step 4: Calculate the difference between the standardized evapotranspiration stress index between adjacent preset time periods within the period to be evaluated, and use the mean and standard deviation of the difference between adjacent preset time periods within the period to be evaluated over many years of history to standardize the difference of the standardized evapotranspiration stress index, and obtain the change in the standardized evapotranspiration stress index. Step 5: Scan the changes in the standardized evapotranspiration stress index for each preset time period according to the time series. When the changes in the standardized evapotranspiration stress index of adjacent preset time periods within the continuous period to be evaluated are lower than the percentile threshold at the beginning of drought, it is determined that the spatial unit has entered the rapid intensification stage, thereby identifying candidate drought events. Step 6: When a candidate drought event meets the following conditions, the duration of the drought event is greater than the percentile threshold of the initial stage of drought, the standardized evapotranspiration stress index at the end of the rapid intensification stage is lower than the 20th percentile threshold of the historical distribution of the multi-year standardized evapotranspiration stress index of spatial unit i, and the average change in the standardized evapotranspiration stress index during the duration of the candidate drought event is lower than the percentile threshold of conventional drought, the candidate drought event is determined as a valid drought event. Step 7: Statistically calculate the annual average frequency, average occurrence rate, and severity of all effective drought events within each spatial unit. After normalizing the annual average frequency, average occurrence rate, and severity, perform weighted calculations according to preset weights to obtain the comprehensive ecological drought risk index.

[0011] As a preferred technical solution of the present invention, in step 1, the actual evapotranspiration data and potential evapotranspiration data are subjected to quality screening and masking processing, including: retaining the image data generated by the main algorithm in the MODIS land product quality control identifier, removing low-quality image data data that are polluted by clouds or have abnormal detection, and masking out missing values ​​and abnormal pixels less than or equal to zero in the potential evapotranspiration data.

[0012] As a preferred embodiment of the present invention, in step 2, the specific calculation formula for calculating the evaporation stress ratio (ESR) of each spatial unit within the target spatial region is as follows:

[0013] in, Representing spatial units In the Year Evaporation stress ratio over a preset time period Representing spatial units In the Year Actual evapotranspiration data that has undergone quality screening within a preset time period. Representing spatial units In the Year Potential evapotranspiration data that has been quality-screened for a preset time period.

[0014] Furthermore, in step 3, the specific calculation formula for the standardized evapotranspiration stress index is as follows:

[0015] in, This represents the standardized evapotranspiration stress index. Representing spatial units In the historical multi-year sequence, the first The average evaporation stress ratio over a preset time period Representing spatial units In the historical multi-year sequence, the first Evaporation stress ratio standard deviation for a preset time period.

[0016] In step 4, the specific formula for calculating the change in the standardized evapotranspiration stress index is as follows:

[0017]

[0018] in, This represents the difference in the standardized evapotranspiration stress index between two adjacent preset time periods. This represents the change in the standardized evapotranspiration stress index. Representing spatial units In the historical sequence of many years, from the first One to the first The mean of the differences over a preset time period. Representing spatial units In the historical sequence of many years, from the first One to the first The standard deviation of the differences over a preset time period.

[0019] As a preferred technical solution of the present invention, in step 5, if during the identification of candidate drought events, a brief mitigation process occurs where the change in the standardized evapotranspiration stress index is higher than the percentile threshold at the beginning of the drought, and the adjacent preset time period in the next evaluation period is lower than the percentile threshold at the beginning of the drought again, and the standardized evapotranspiration stress index after the mitigation process is lower than the standardized evapotranspiration stress index before the mitigation process, then the candidate drought event is kept uninterrupted; if two or more brief mitigation processes occur consecutively, then the identification of the rapid enhancement stage is terminated.

[0020] In practical use, the adjacent preset time period within the evaluation period can specifically be eight days. Furthermore, in step 7, the specific formula for calculating the annual average frequency is:

[0021] in, Representing spatial units The average annual frequency, Representing spatial units The total number of valid drought events identified within the total time span. This indicates the total number of years spanned.

[0022] As a preferred embodiment of the present invention, the specific formula for calculating the average occurrence rate in step 7 is as follows:

[0023] in, Representing spatial units No. The event occurrence rate of a sub-effective drought event and These represent the preset start and end time periods of the rapid intensification phase of an effective drought event, respectively. This represents the total number of standardized evapotranspiration stress index changes included in the effective drought event. This represents the absolute value of the change in the standardized evapotranspiration stress index during an effective drought event; for spatial units. The average occurrence rate is obtained by averaging the occurrence rates of all effective drought events over multiple years.

[0024] As a preferred embodiment of the present invention, the specific formula for calculating the severity in step 7 is as follows:

[0025] in, Representing spatial units No. The severity of a single effective drought event. Indicates the first Standardized evapotranspiration stress index for each preset time period during the rapid intensification phase of the effective drought event. Represents the absolute value of the standardized evapotranspiration stress index over a preset time period; for spatial units The severity intensity is obtained by averaging the individual severity intensities of all effective drought events over multiple years.

[0026] In step 7, the specific formula for calculating the comprehensive risk index of ecological drought is as follows:

[0027] in, Representing spatial units The comprehensive risk index of sudden ecological drought. This represents the annual average frequency after normalization. This represents the average occurrence rate after normalization. This indicates the severity level after normalization. The preset weight representing the average annual frequency. The preset weights representing the average occurrence rate The preset weights representing the severity level, and The preset weights are determined using equal weight assignment, entropy weighting, or analytic hierarchy process.

[0028] As a preferred embodiment of the present invention, step 7, after obtaining the comprehensive risk index of sudden ecological drought, further includes a corrected assessment step based on vegetation resilience: Acquire actual and potential evapotranspiration data within a preset time period after the end of all effective drought events in the target spatial area, and calculate the standardized evapotranspiration stress index during the recovery period; The duration of vegetation recovery for each effective drought event was calculated based on the standardized evapotranspiration stress index during the recovery period. Calculate the average vegetation recovery time for all effective drought events within a spatial cell, and construct a vegetation vulnerability penalty coefficient; By using the vegetation vulnerability penalty coefficient to nonlinearly adjust the comprehensive risk index of sudden ecological drought, a risk assessment level of limited ecological restoration is obtained.

[0029] As a preferred embodiment of the present invention, the calculation process for the single vegetation restoration time and the vegetation vulnerability penalty coefficient is as follows: For the first within the spatial unit For a single effective drought event, when the standardized evapotranspiration stress index of the recovery period is greater than or equal to zero twice in a series of preset time periods after the effective drought event ends, the time span from the end of the effective drought event to the first time it is greater than or equal to zero is recorded as the single vegetation recovery duration. The specific formula for calculating the average vegetation recovery time is as follows:

[0030]

[0031] in, Representing spatial units The average vegetation recovery time Representing spatial units The total number of effective drought events within the region. Representing spatial units No. The duration of vegetation recovery during a single effective drought event; the specific formula for calculating the vegetation vulnerability penalty coefficient is as follows:

[0032]

[0033] in, Representing spatial units The vegetation vulnerability penalty coefficient, This represents the preset upper limit constant for penalty intensity. This represents the preset curve slope factor. The time constant represents the baseline for vegetation restoration; the specific formula for calculating the risk assessment level of limited ecological restoration is as follows:

[0034]

[0035] in, Representing spatial units The risk assessment level of ecological restoration limitations Representing spatial units The comprehensive risk index of sudden ecological drought.

[0036] Compared with the prior art, the present invention has the following beneficial effects: This invention constructs a surge drought identification framework based on a standardized evapotranspiration stress index and its variation by coupling actual and potential evapotranspiration of vegetation. This framework directly reflects the vegetation's response to water deficit and the land-atmosphere coupling process, overcoming the shortcomings of traditional methods that rely on soil moisture or single rate indicators, resulting in identification lag and insufficient ecological interpretation. The evapotranspiration process is located at a critical interface of the soil-vegetation-atmosphere continuum, comprehensively reflecting the combined effects of multiple factors such as high temperature, low humidity, enhanced radiation, wind speed changes, and precipitation deficit on vegetation water use, making the detection of surge drought events more sensitive and accurate. By standardizing the evapotranspiration stress ratio based on the multi-year mean and standard deviation of the same period in history, background differences between different climatic zones, vegetation types, and phenological periods are eliminated, achieving consistent identification and comparison of surge drought events on a global scale. This significantly improves inter-regional comparability and is applicable to drought monitoring across large areas and multiple ecosystem types.

[0037] This invention employs a rapid enhancement phase identification method based on changes in the standardized evapotranspiration stress index, and applies multiple constraints including duration, end-stage stress state, and overall rate of change. This effectively distinguishes between sudden drought and evapotranspiration fluctuations caused by short-term heat and dryness fluctuations or normal vegetation phenological changes, significantly reducing the misjudgment rate and improving the reliability of event identification. It comprehensively characterizes event features from three dimensions: frequency of occurrence, development rate, and stress intensity, and generates an ecological sudden drought risk index through a three-dimensional weighted comprehensive assessment, avoiding the one-sidedness of traditional single-indicator evaluations.

[0038] This invention further introduces a modified assessment based on vegetation resilience. By calculating the post-disaster vegetation recovery time and constructing a nonlinear penalty coefficient, it characterizes the differentiated recovery capabilities of different vegetation types after encountering extreme drought, enabling accurate identification of hidden high-risk plots and providing a more scientific basis for decision-making in ecological restoration and shelterbelt construction. Attached Figure Description

[0039] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0040] Figure 1 This is a flowchart illustrating the overall technical process of the method of this invention. Detailed Implementation

[0041] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the embodiments of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0042] The following is in conjunction with the appendix Figure 1 The embodiments of the present invention will be described in detail below.

[0043] Example 1: This invention discloses a method for identifying and assessing the risk of sudden drought by coupling vegetation evapotranspiration dynamics. To achieve the purpose of this invention, the following technical solution is adopted, which mainly includes two tasks: a global high-precision resolution long-term series sudden drought event identification framework and a comprehensive assessment of the global regional sudden drought risk level.

[0044] The framework for identifying sudden drought events comprises three main modules. The implementation of this framework is based on long-term, high-precision global actual and potential evapotranspiration datasets acquired by the Moderate Resolution Imaging Spectroradiometer (MODIS) aboard the Terra and Aqua Earth observation satellites. It uses the occurrence conditions of sudden drought events as the basis for judgment and captures event characteristics such as occurrence speed and intensity.

[0045] In practice, during the successive event identification process, high-precision global evapotranspiration and potential evapotranspiration data are first subjected to rigorous quality screening and preprocessing to calculate the standardized evapotranspiration stress index for long-term series. This dataset can reflect the changes in evapotranspiration stress of each spatial unit relative to the multi-year average climate state for each preset time period.

[0046] In this embodiment, the preset time period is specifically an 8-day composite period, and the spatial unit is specifically a pixel of a remote sensing image. Following the natural process of a sudden drought event, multiple occurrence conditions are aligned, and each pixel is judged to determine whether the initial conditions, the rate of decline during the outbreak, and the shortest duration under drought conditions meet the preset time period parameters, thereby identifying candidate sudden drought events.

[0047] Furthermore, a global assessment is needed based on the total duration of each candidate event and the average change during the event period. The final output records the drought events that simultaneously meet both temporal and global conditions. Finally, feature calculations are performed on the selected drought events, statistically analyzing their frequency, rate of occurrence, and intensity. The second major module is the comprehensive assessment of ecological drought risk levels. Based on the three-dimensional characteristics of drought events, a combined assessment using weighted scoring and segmented grading is conducted to achieve a comprehensive evaluation, resulting in a regional ecological drought risk level distribution map that reveals high-risk areas for ecological drought globally.

[0048] Detailed technical process as follows Figure 1 As shown.

[0049] In practical implementation, this embodiment includes the following steps: Step 1. Acquisition, quality screening, and preprocessing of high-precision global data on actual vegetation evapotranspiration and potential evapotranspiration of vegetation environment; To characterize the dynamics of vegetation evapotranspiration and the potential evapotranspiration capacity of the environment in which vegetation physiological growth activities occur, high-precision evapotranspiration and potential evapotranspiration datasets with long-term time series were selected as the basic input data. These datasets consist of actual evapotranspiration (ET) and potential evapotranspiration (PET) from MOD16A2GF. The datasets are global-scale long-term time series products with a spatial resolution of 500m and a temporal resolution of 8 days, which can meet the needs of large-scale, long-term vegetation evapotranspiration stress monitoring.

[0050] In the data preprocessing process, ET, PET, and ET_QC quality control layers for each 8-day period were first obtained, and ET and PET were scaled according to the product scaling factor. Then, the pixels were quality-screened based on the ET_QC layer: pixels identified by the MODIS Land Product Quality Control (MODLAND_QC, a standardized pixel-level quality assessment field for MODIS land products) that were generated by the main algorithm and were of good quality were prioritized; low-quality pixels affected by severe cloud contamination, dead pixels, main algorithm failure, or those that could not be reliably inverted were removed; pixels with filled, outlier, or null ET or PET values, or PET values ​​less than or equal to 0, were masked. Composite periods of less than 8 days at the end of the year were uniformly treated as the last 8-day composite period of that year and numbered accordingly. After quality control, long-term series data of ET and PET for each pixel, year, and 8-day composite period were obtained. At the same time, based on this historical data, the mean evaporative stress ratio, standard deviation of evaporative stress ratio, mean difference, and standard deviation of difference are calculated and stored for each 8-day composite period for each pixel, forming a historical benchmark database.

[0051] Step 2. Calculate the evaporation stress ratio of the identified parameters; Based on data that has undergone rigorous quality control, a framework for identifying sudden droughts was implemented. First, the evaporative stress ratio (ESR) was calculated using the ratio of actual evapotranspiration (ET) to potential evapotranspiration (PET). This value characterizes the degree to which actual vegetation evapotranspiration meets the potential evapotranspiration demand of the environment. A higher ESR (closer to 1) indicates that land surface water and vegetation water supply can adequately meet atmospheric evapotranspiration demand, while a lower ESR (closer to 0) indicates that atmospheric evapotranspiration demand is difficult to meet with surface water supply, and the ecosystem experiences stronger evapotranspiration stress. The calculation formula is as follows:

[0052]

[0053] in For the first The pixel in the first Year Evaporation stress ratio over an 8-day synthetic period This represents the actual evaporation after quality screening for the corresponding time period. This represents the potential evapotranspiration after quality screening for the corresponding time period. Abnormal ESR values ​​occurring during the calculation process are eliminated or constrained based on ET, PET quality indicators, and reasonable physical ranges.

[0054] Step 3. Calculate the standardized evapotranspiration stress index of the identified parameters; Due to significant differences in evapotranspiration background across different climate zones, vegetation types, and phenological periods, direct use of ESR for cross-regional comparisons is difficult. Therefore, this scheme performs multi-year standardization on each pixel and each 8-day composite period within a year to obtain the Standardized Evaporative Stress Ratio (SESR). The calculation formula is as follows:

[0055]

[0056] in This represents the standardized evapotranspiration stress index. Indicates the first The first pixel in the historical time series The average ESR over an 8-day composite period This represents the corresponding standard deviation. This calculation can unify the evapotranspiration stress states of different regions and seasons to a relatively multi-year climatological anomaly level, thereby improving the comparability of surge drought identification results across global scales and different ecosystems.

[0057] Step 4. Calculate the standardized evapotranspiration stress index change of the identified parameters; The change in the standardized evapotranspiration stress index (SESR) is the last and most important key identification parameter. To characterize the rapid onset and intensification of a flash drought event, this scheme further calculates the change in SESR between adjacent 8-day composite periods, i.e. This data is also standardized to characterize the rapid changes in the evapotranspiration stress index across different vegetation types and phenological stages. The calculation formula is as follows:

[0058]

[0059]

[0060] in This indicates the change in SESR between two adjacent 8-day composite periods. Indicates the first The pixel in the first Mid-year Evapotranspiration stress index for each time period; This represents the change in the standardized evapotranspiration stress index. Indicates the first In the historical sequence of many years, from the first pixel... One to the first The average change in SESR over an 8-day composite period This represents the corresponding standard deviation. The smaller the value, the faster the SESR decreases and the more pronounced the evapotranspiration stress, making it a key indicator for identifying the rapid onset phase of a sudden drought.

[0061] Step 5. Determining the critical rapid intensification phase of a sudden drought event; After obtaining the SESR and the changes in the standardized evapotranspiration stress index, the time series is scanned year by year according to the pixel to identify possible rapid intensification phases of flash drought. When the change in the standardized evapotranspiration stress index of a certain adjacent period is lower than the 10th percentile threshold of the historical distribution of the corresponding period for that pixel, the period is considered to meet the conditions for rapid intensification of evapotranspiration stress and can be used as the starting point for judging candidate flash drought events.

[0062] During the ongoing identification of candidate events, if the change in the Standardized Evapotranspiration Stress Index (SESR) in subsequent adjacent time periods remains below the 10th percentile threshold, the event is considered to continue developing. If a brief mitigation occurs where the change in the SESR exceeds the 10th percentile threshold, the candidate event is allowed to continue temporarily, but the rapid decline condition must be met again in the next adjacent time period, and the SESR after mitigation should remain below the pre-mitigation level. If two or more mitigation periods occur consecutively, or if the subsequent SESR no longer shows an increasing trend of evapotranspiration stress, the identification of the rapid enhancement phase of the candidate event is terminated.

[0063] Step 6. Aridity constraints; To avoid misjudging short-term fluctuations in dry heat or normal phenological changes in vegetation as sudden drought events, this scheme introduces drought state constraints and duration constraints based on the determination of the rapid intensification stage.

[0064] Specifically, a candidate flash flood event must simultaneously meet the following conditions: First, the event must contain at least five consecutive SESR changes, meaning the event duration must be at least five 8-day composite periods. Second, the SESR at the end of the rapid intensification phase should be lower than the 20th percentile threshold of the historical SESR distribution for that pixel's corresponding time period, ensuring that the event eventually enters a significant evapotranspiration stress state. Third, the average change in the standardized evapotranspiration stress index during the entire candidate event period should be lower than the 15th percentile threshold of the historical standardized evapotranspiration stress index change distribution for that pixel's corresponding time period, ensuring that the event as a whole exhibits rapid development characteristics. Only candidate events that simultaneously meet the constraints of rapid decline, end-of-phase drought state, duration, and overall rapid development are identified as valid flash flood events.

[0065] Step 7. Statistics and Risk Assessment: Step 7.1 Output the successive drought events pixel by pixel year by year; After completing the above determination, each valid drought event identified in each pixel and each year is recorded sequentially. Each drought event includes at least the following information: the pixel where the event occurred, the year of occurrence, the start time, the end time, the duration, the SESR sequence during the event, the standardized evapotranspiration stress index change sequence, the final SESR value, and the event's average standardized evapotranspiration stress index change.

[0066] When multiple independent flash drought events occur in the same pixel within the same year, they are recorded and numbered chronologically. Two candidate events are considered independent if there is a clear recovery phase between them, meaning the SESR (Sequencing and Reduction Regulatory Situation) no longer maintains a drought state or evapotranspiration stress is significantly alleviated. For example, an event can be considered terminated when the SESR rises above the 30th percentile of the historical SESR for that pixel's corresponding time period, and the duration is no less than two 8-day composite periods; subsequent events are then considered independent new events. This process ultimately creates a global list of flash drought events on a pixel-by-pixel, year-by-year, and event-by-event basis, providing fundamental data for subsequent frequency, rate, and intensity statistics.

[0067] Step 7.2 Statistical analysis of the annual average frequency characteristics of events; Based on the list of successive drought events obtained in step 7.1, the total number of drought events occurring in each pixel during the entire study period was counted, and the annual average frequency was calculated. The calculation method is as follows:

[0068]

[0069] in Indicates the first The average annual frequency of sudden droughts per pixel This indicates the total number of drought events identified in that pixel during the study period. This represents the total number of years studied. This indicator is used to characterize the long-term frequency of flash drought events in a region. The higher the frequency, the greater the likelihood that the region will be repeatedly affected by flash droughts, and the higher the long-term exposure risk to the ecosystem.

[0070] Step 7.3 Statistical analysis of the occurrence rate characteristics of events; To characterize the development speed of sudden drought events, this scheme calculates the event occurrence rate based on the sequence of standardized evapotranspiration stress index changes during the rapid intensification phase of each event. For the th The occurrence rate of this sudden drought event can be expressed as:

[0071]

[0072] in Indicates the first One pixel The rate of occurrence of flash drought events and These represent the beginning and end of the rapid escalation phase of the event, respectively. Indicates the number of SESR changes contained in the event. This represents the absolute value of the change in the standardized evapotranspiration stress index during the event. Subsequently, the occurrence rates of all events over multiple years in the same pixel are averaged to obtain the multi-year average flash drought rate for that pixel. A larger index indicates a faster increase in evapotranspiration stress within a short period, and a more intense flash drought.

[0073] Step 7.4 Statistical analysis of the intensity characteristics of the event; To characterize the severity of vegetation evapotranspiration stress caused by a sudden drought event, this scheme uses the cumulative SESR deficit during the rapid intensification phase of the event to represent the event intensity.

[0074]

[0075] in Indicates the first One pixel The severity of this sudden drought event, This represents the standardized evapotranspiration stress index for each period during the rapid intensification phase of the event. The average severity intensity of a pixel is obtained by averaging the severity intensities of all drought events over multiple years, and is used to characterize the potential impact of regional droughts on vegetation ecosystems.

[0076] Step 7.5 Three-dimensional comprehensive evaluation of the event; After obtaining three dimensions of indicators—annual average frequency, occurrence rate, and severity—this scheme normalizes each of the three indicators to convert them to a unified dimensional range. This embodiment employs the max-min normalization method for any indicator. Its normalized value ,in and These represent the minimum and maximum values ​​of the indicator across all spatial units within the study area, respectively. After normalization, all indicator values ​​are mapped to the [0,1] interval. Subsequently, a comprehensive calculation is performed according to preset weights to obtain the ecological drought risk index:

[0077]

[0078] in Indicates the first The comprehensive risk index of sudden ecological drought per pixel. , , These represent the normalized annual average frequency, average occurrence rate, and average severity of sudden droughts, respectively. , , These represent the weights of the three indicators, and the sum of the weights equals 1.

[0079] In practical implementation, equal-weighted assignment can be adopted (i.e. The weights can also be determined based on expert experience, entropy weighting, or the analytic hierarchy process (AHP). After the comprehensive risk index is calculated, global pixels are divided into low-risk, lower-risk, medium-risk, higher-risk, and high-risk levels according to the natural breakpoint method, quantile method, or preset threshold method, ultimately generating a spatial distribution map of global ecological drought risk levels. The annual average frequency reflects the long-term exposure level of a region, the occurrence rate reflects the speed of drought outbreaks, and the intensity reflects the degree of vegetation evapotranspiration stress. Through three-dimensional comprehensive assessment, risk bias caused by single-indicator evaluation can be avoided, and the comprehensive risk level of ecological drought in different regions of the world can be revealed more comprehensively.

[0080] Step 7.6 Output the results; After completing the above sub-steps, the system outputs global pixel-by-pixel drought event identification results, an annual list of drought events, a distribution map of the annual average frequency of droughts, a distribution map of drought occurrence rates, a distribution map of drought severity, and a distribution map of the comprehensive ecological drought risk level. This enables a method for identifying and assessing drought events based on vegetation evapotranspiration dynamics, providing technical support for global or regional-scale agricultural drought monitoring, forest degradation risk identification, ecosystem function assessment, and disaster early warning.

[0081] Example 2: This example is an optimization based on Example 1, proposing a risk assessment correction scheme based on vegetation resilience characteristics.

[0082] The calculated comprehensive ecological drought risk index focuses on the degree of exposure to external environmental stress during the disaster occurrence phase, encompassing parameters such as frequency, rate, and intensity. Vegetated ecosystems exhibit differentiated inherent resilience in the face of extreme water loss. Some drought-resistant spatial units, after experiencing intense evapotranspiration stress, show rapid rebound in stomatal conductance and leaf water potential with the arrival of precipitation pulses. Other spatial units, constrained by deep soil moisture deficits or xylem embolism, exhibit prolonged lag growth inhibition after the effective drought event ends, resulting in carbon starvation or irreversible canopy damage. The length of the post-disaster recovery cycle constitutes a core variable in the long-term vulnerability of ecosystems. Introducing a time-span-based resilience assessment can accurately map the ecological memory effect in the land-atmosphere coupling system, ensuring that the final risk level matches the actual vegetation degradation trajectory.

[0083] This embodiment performs resilience assessment based on the end time of an effective drought event. Spatial units are defined. In the The preset time period for the end of the effective drought event is: Time series scanner The time period is then slid forward to extract actual and potential evapotranspiration data for the corresponding intervals. Following the calculation process from steps 1 to 4, a standardized evapotranspiration stress index sequence for the recovery period is obtained. This sequence reflects the dynamic process of actual water consumption reverting to the historical climatic mean after vegetation leaves the rapid enhancement phase and exceeds the drought termination threshold.

[0084] The recovery criteria require the standardized evapotranspiration stress index to be greater than or equal to zero twice consecutively during the recovery period. Zero represents the average level of the same climatological state over many years in the standardized distribution; meeting this criterion twice consecutively aims to exclude false recovery phenomena caused by a single abnormal rainfall pulse. The first preset time period to meet this criterion is denoted as... Duration of single vegetation restoration From the formula The calculation yielded the result.

[0085] There are two specific natural conditions that could lead to scan overflow. First, a spatial cell might encounter another independent climate disturbance during its recovery period, triggering another effective drought event. Second, the recovery process might coincide with the end of the vegetation growing season, the beginning of winter dormancy or leaf fall, in which case the significant decrease in actual evapotranspiration data would be phenologically driven rather than drought-stressed. For these conditions, a mandatory truncation constraint is set: a preset maximum search duration constant is used. The value is set to 12 preset time periods. If from... rise If the condition of two consecutive zero values ​​is not met within a certain time period, forced recording will occur. The characterization system encountered extreme recovery lag; dynamic phenological mask parameters were introduced to determine the lower limits of surface temperature and radiation flux. When a spatial unit was detected to have entered the non-growing season, the current state was frozen. The scanning progress excludes preset time periods during dormancy, and the time span continues to accumulate until the phenological greening node of the following spring.

[0086] The arithmetic mean of the single vegetation recovery time corresponding to all effective drought events within the spatial unit was calculated to obtain the average vegetation recovery time, which represents the long-term resilience characteristics of the spatial unit. .by A vegetation vulnerability penalty coefficient is constructed by using independent variables to drive a nonlinear logistic regression function. The logistic function's morphological characteristics conform to the biological limit response law: the function base is 1, ensuring that spatial units with extremely fast recovery speeds do not change their original comprehensive ecological drought risk index; as the average vegetation recovery time increases, the penalty coefficient rises non-linearly in an "S" shape, with a gentle rise in the early stage, allowing for natural recovery fluctuations within a certain range; it crosses the vegetation recovery baseline time constant. After the corresponding inflection point, the penalty intensity increases sharply, indicating that the system is on the verge of collapse. In the parameter settings, The value is assigned to 0.5, which determines the maximum upper limit of the limited risk amplification to be 1.5 times the base risk value; A value of 0.8 is assigned to dominate the transition rate in the sensitive region; A value of 4 maps to the baseline buffer period under normal conditions. Using the generated... Comprehensive risk index of sudden ecological drought Perform a multiplication operation to output an ecological restoration limitation risk assessment level that takes into account both stress intensity and ecological memory effect. .

[0087] To verify the effect of introducing resilience factor correction in this embodiment, a retrospective simulation over a ten-year period (2011-2020) was conducted on four heterogeneous spatial units (pixels A, B, C, and D) within the mid-latitude forest-steppe transition zone of the Northern Hemisphere. The four pixels had similar initial ecological drought risk indices output based on the method in Example 1, all falling within the range of 0.65 to 0.70. According to the traditional single evaluation system, these would be uniformly classified as high-risk areas. The underlying data sequences of each pixel were extracted, and the scanning and nonlinear penalty algorithm of this embodiment was executed.

[0088] Table 1. Comparative Assessment of Ecological Disaster Risks Incorporating Vegetation Restoration Capacity;

[0089]

[0090] Pixels A and D, leveraging their deep root systems or strong hydraulic regulation mechanisms, were able to rapidly reduce evapotranspiration stress to zero after each effective drought event. The average vegetation recovery time was only 2.0 and 2.2 preset time intervals, respectively, far lower than the baseline time constant. The logistic function generated... Approaching the lower limit, the final calculated risk assessment level for limited ecological restoration The leaf area index (LAI) increased slightly to 0.71 and 0.74, maintaining the original risk assessment range. Actual optical remote sensing verification confirmed that neither pixel showed an interannual decline in LAI over the past decade, indicating a intact ecological structure.

[0091] Pixel C exhibits system vulnerability. Its parameters during the disaster are highly similar to those of pixel A, but due to prior habitat degradation and pest / disease disturbances, root zone water absorption efficiency is severely depleted. Time-series scanners recorded extremely difficult recovery after each drought, with the final sudden drought triggering a 12-period cutoff mechanism. The average recovery time was as high as 8.4, placing it in the terminal saturation zone of the S-curve, with a penalty coefficient reaching 1.48. Risks previously masked within the general high-risk range were amplified to 1.02, jumping to an area of ​​extremely high degradation risk. This risk level differentiation precisely corresponds to the vegetation death patch succession already present in the optical imagery of this pixel.

[0092] Pixel B represents a sensitive ecological niche in a critical state. The average recovery time is... The system, which patrolled the vicinity and detected a penalty coefficient of 1.27, identified a technical signal that the area was on the verge of losing its resilience, raising its risk level to 0.83 and serving as a preventative early warning.

[0093] This embodiment combines pure time-series signal extraction with unidirectional logistic regression, resulting in linearly increasing time complexity, which can be integrated into a global 500m resolution parallel processing matrix. The ecological restoration limitation risk assessment level serves as the final supplementary output, establishing a closed-loop link from water-air imbalance during disasters to post-disaster biophysiological stress. This directly supports the identification of hidden high-risk areas with poor endogenous resilience and urgent need for artificial water resource allocation intervention during large-scale ecological planning and shelterbelt construction, by stripping away surface drought intensity.

[0094] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. It should be noted that any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics, characterized in that, Includes the following steps: Step 1: Obtain actual evapotranspiration data and potential evapotranspiration data for multiple consecutive preset time periods within the target spatial area, and perform quality screening and masking processing on the actual evapotranspiration data and potential evapotranspiration data; Step 2: Based on the filtered actual evapotranspiration data and potential evapotranspiration data, calculate the evaporation stress ratio of each spatial unit within the target spatial region; Step 3: Obtain the mean and standard deviation of the evaporation stress ratio for the target spatial region over many years corresponding to a preset time period. Then, use the mean and standard deviation of the evaporation stress ratio to standardize the calculated evaporation stress ratio and obtain the standardized evaporation stress index. Step 4: Calculate the difference between the standardized evapotranspiration stress index between adjacent preset time periods within the period to be evaluated, and use the mean and standard deviation of the difference between adjacent preset time periods within the period to be evaluated over many years of history to standardize the difference of the standardized evapotranspiration stress index, and obtain the change in the standardized evapotranspiration stress index. Step 5: Scan the changes in the standardized evapotranspiration stress index for each preset time period according to the time series. When the changes in the standardized evapotranspiration stress index of adjacent preset time periods within the continuous period to be evaluated are lower than the percentile threshold at the beginning of drought, it is determined that the spatial unit has entered the rapid intensification stage, thereby identifying candidate drought events. Step 6: When a candidate drought event meets the following conditions, the duration of the drought event is greater than the percentile threshold of the initial stage of drought, the standardized evapotranspiration stress index at the end of the rapid intensification stage is lower than the 20th percentile threshold of the historical distribution of the multi-year standardized evapotranspiration stress index of spatial unit i, and the average change in the standardized evapotranspiration stress index during the duration of the candidate drought event is lower than the percentile threshold of conventional drought, the candidate drought event is determined as a valid drought event. Step 7: Statistically calculate the annual average frequency, average occurrence rate, and severity of all effective drought events within each spatial unit. After normalizing the annual average frequency, average occurrence rate, and severity, perform weighted calculations according to preset weights to obtain the comprehensive ecological drought risk index.

2. The method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics according to claim 1, characterized in that: In step 1, quality screening and masking are performed on the actual evapotranspiration data and potential evapotranspiration data, including: retaining the image data generated by the main algorithm in the MODIS land product quality control identifier, removing low-quality image data data that are polluted by clouds or have abnormal detection, and masking out missing values ​​and abnormal pixels less than or equal to zero in the potential evapotranspiration data.

3. The method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics according to claim 1, characterized in that: In step 2, the specific formula for calculating the evaporation stress ratio of each spatial unit within the target spatial region is as follows: in, Representing spatial units In the Year Evaporation stress ratio over a preset time period Representing spatial units In the Year Actual evapotranspiration data that has undergone quality screening within a preset time period. Representing spatial units In the Year Potential evapotranspiration data that has been quality-screened for a preset time period.

4. The method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics according to claim 3, characterized in that: In step 3, the specific formula for calculating the standardized evapotranspiration stress index is as follows: in, This represents the standardized evapotranspiration stress index. Representing spatial units In the historical multi-year sequence, the first The average evaporation stress ratio over a preset time period Representing spatial units In the historical multi-year sequence, the first Evaporation stress ratio standard deviation for a preset time period.

5. The method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics according to claim 4, characterized in that: In step 4, the specific formula for calculating the change in the standardized evapotranspiration stress index is as follows: in, This represents the difference in the standardized evapotranspiration stress index between two adjacent preset time periods. This represents the change in the standardized evapotranspiration stress index. Representing spatial units In the historical sequence of many years, from the first One to the first The mean of the differences over a preset time period. Representing spatial units In the historical sequence of many years, from the first One to the first The standard deviation of the differences over a preset time period.

6. The method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics according to claim 1, characterized in that: In step 5, if during the candidate drought event identification process, there is a brief mitigation process in which the change in the standardized evapotranspiration stress index is higher than the percentile threshold at the beginning of the drought, and the adjacent preset time period in the next assessment period is lower than the percentile threshold at the beginning of the drought again, and the standardized evapotranspiration stress index after the mitigation process is lower than the standardized evapotranspiration stress index before the mitigation process, then the candidate drought event is kept uninterrupted. If two or more brief mitigation processes occur consecutively, the rapid enhancement phase identification will be terminated.

7. The method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics according to claim 1, characterized in that: In step 7, the specific formula for calculating the annual average frequency is as follows: in, Representing spatial units The average annual frequency, Representing spatial units The total number of valid drought events identified within the total time span. This indicates the total number of years spanned.

8. The method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics according to claim 1, characterized in that: In step 7, the specific formula for calculating the average occurrence rate is as follows: in, Representing spatial units No. The event occurrence rate of a sub-effective drought event and These represent the preset start and end time periods of the rapid intensification phase of an effective drought event, respectively. This represents the total number of standardized evapotranspiration stress index changes included in the effective drought event. This represents the absolute value of the change in the standardized evapotranspiration stress index during an effective drought event; for spatial units. The average occurrence rate is obtained by averaging the occurrence rates of all effective drought events over multiple years.

9. The method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics according to claim 8, characterized in that: In step 7, the specific formula for calculating the severity is as follows: in, Representing spatial units No. The severity of a single effective drought event. Indicates the first Standardized evapotranspiration stress index for each preset time period during the rapid intensification phase of the effective drought event. Represents the absolute value of the standardized evapotranspiration stress index over a preset time period; for spatial units The severity intensity is obtained by averaging the individual severity intensities of all effective drought events over multiple years.

10. The method for identifying and assessing the risk of sudden drought coupled with vegetation evapotranspiration dynamics according to claim 1, characterized in that, Step 7, after obtaining the comprehensive risk index of ecological drought, also includes a corrected assessment step based on vegetation resilience: Acquire actual and potential evapotranspiration data within a preset time period after the end of all effective drought events in the target spatial area, and calculate the standardized evapotranspiration stress index during the recovery period; The duration of vegetation recovery for each effective drought event was calculated based on the standardized evapotranspiration stress index during the recovery period. Calculate the average vegetation recovery time for all effective drought events within a spatial cell, and construct a vegetation vulnerability penalty coefficient; By using the vegetation vulnerability penalty coefficient to nonlinearly adjust the comprehensive risk index of sudden ecological drought, a risk assessment level of limited ecological restoration is obtained.

Images