Method for assessing vegetation loss risk under the stress of rapid transition from drought to flood, and device for assessing vegetation loss risk under the stress of combined dry-wet events.
The method addresses the underestimation of vegetation loss risk during drought-to-flood transitions by using precipitation and NDVI data to model composite events, enhancing the accuracy of risk assessment and ecological management.
Patent Information
- Application Number
- JP2026188435
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2025-08-19
- Filing Date
- 2026-08-14
- Publication Date
- 2026-11-19
- Estimated Expiration
- 2046-08-14
AI Technical Summary
Conventional disaster risk assessment methods fail to accurately quantify the synergistic effects of abrupt transitions from drought to flood on vegetation, neglecting dynamic changes in vegetation's adaptive capacity and exposure levels, leading to underestimation of ecological loss risk.
A method using seasonal precipitation data and NDVI data to identify dry-wet composite events, calculate their probabilities, and assess vegetation loss risk through a Bayesian framework combined with the Copula function, considering exposure levels to construct a combined distribution of composite events.
Enhances the accuracy of vegetation loss risk assessment by modeling actual disaster chains, identifying high-risk areas, and improving ecological management strategies under climate change.
Smart Images

Figure 0007953965000001_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application belongs to the field of ecological disaster risk assessment, and more specifically, relates to a method for assessing vegetation loss risk under the stress of abrupt transition from drought to flood. BACKGROUND ART
[0002] With the intensification of global climate change, the frequency and intensity of dry-wet compound events such as the abrupt transition from drought to flood continue to increase, bringing severe stress and damage to ecosystems, especially vegetation.
[0003] However, most of the conventional disaster risk assessment methods focus on a single disaster (e.g., drought or flood) and have many limitations. For example, the quantification of the synergistic effect of dry-wet compound events is insufficient, the superposition effect of hydrological stress during the dry-wet transition process is ignored, which may lead to underestimation of the ecological loss risk; in addition, the existing vulnerability assessment is based on static indicators, the dynamic response of the disaster-affected body under compound events is not considered, and the single disaster model cannot capture the chain effect between various events. Due to these problems, the conventional disaster risk assessment methods cannot accurately assess the vegetation loss risk.
[0004] Accordingly, there is an urgent need for a method for assessing vegetation loss risk under the stress of abrupt transition from drought to flood, so as to more accurately assess the vegetation loss risk and provide a scientific basis for ecological protection and disaster management. SUMMARY OF THE INVENTION PROBLEMS TO BE SOLVED BY THE INVENTION
[0005] In view of the shortcomings of the prior art, the purpose of the present application is to provide a method for assessing vegetation loss risk under the stress of abrupt transition from drought to flood, so as to more accurately assess the vegetation loss risk and provide a scientific basis for ecological protection and disaster management. MEANS FOR SOLVING THE PROBLEMS
[0006] To achieve the above objective, in a first aspect, the present invention provides a method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood. Step S10 involves acquiring monthly precipitation data and NDVI data for the target area, processing them, and obtaining gridded seasonal-scale time series of precipitation data and NDVI data. S20, based on a time series of precipitation data on a seasonal scale, identifies inter-seasonal dry-wet composite events, categorizes them into moderate and severe scenarios, then calculates the probability of inter-seasonal dry-wet composite events occurring under different scenarios using the Copula function, and simultaneously calculates the probability of vegetation loss under stress for inter-seasonal dry-wet composite events under different scenarios by combining the Bayesian framework and the Copula function. S30, a step of calculating the exposure of the vegetation system to a combined dry-wet event using a seasonally averaged normalized vegetation index based on a seasonal time series of NDVI data, S40 includes the step of calculating the risk of vegetation loss under stress from a combined dry and wet event, based on the probability of occurrence and loss calculated in step S20 and the degree of exposure calculated in step S30.
[0007] The present invention's method for evaluating vegetation loss risk under the stress of a rapid transition from drought to flood has the following effects: (1) By constructing a combined distribution of composite events and quantifying the dependence of composite events through composite event probability modeling based on the Copula function, it better fits the characteristics of actual disaster chains compared to the superposition of single disaster probabilities. (2) By prioritizing the exposure level of natural systems compared to exploring risk primarily based on human systems (population, GDP), it becomes easier to identify high-risk vegetation areas (e.g., wetland-grassland transition zones), improving spatial targeting for ecological management, and at the same time providing important reference value for preventing and controlling ecological risks under climate change.
[0008] More preferably, in step S20, the step of identifying dry-wet compound events between adjacent seasons based on the seasonal-scale time series of precipitation data and classifying them into moderate and severe scenarios is specifically as follows: Calculate the SPI index based on the seasonal-scale time series of precipitation data, According to the classification criteria for drought levels, extract drought events and wet events using the SPI index, and classify them into moderate and severe scenarios, Based on the seasons in which drought events and wet events occur, identify dry-wet compound events between adjacent seasons, and the compound events include four of the following: drought-to-wet transition from spring to summer, wet-to-drought transition from spring to summer, continuous drought from spring to summer, continuous wet from spring to summer, drought-to-wet transition from summer to autumn, wet-to-drought transition from summer to autumn, continuous drought from summer to autumn, continuous wet from summer to autumn, drought-to-wet transition from autumn to winter, wet-to-drought transition from autumn to winter, continuous drought from autumn to winter, continuous wet from autumn to winter, drought-to-wet transition from winter to spring, wet-to-drought transition from winter to spring, continuous drought from winter to spring, and continuous wet from winter to spring.
[0009] More preferably, the classification criteria for the drought level are: -2<SPI≦-1.5 corresponds to severe drought, -1.5<SPI≦-1 corresponds to moderate drought, 1<SPI≦1.5 corresponds to moderate flood, and 1.5<SPI≦2 corresponds to severe flood.
[0010] More preferably, in step S20, the step of calculating the occurrence probability of dry-wet compound events between adjacent seasons under different scenarios using the Copula function is specifically as follows: Fit the marginal distributions of the SPI index of adjacent seasons using the normal distribution and the generalized extreme value distribution, and select the optimal marginal distribution using the Akaike Information Criterion, Using the root mean square error and AIC criterion, select the optimal joint distribution function for the SPI time series between adjacent seasons from the Clayton-Copula, Frank-Copula, Gumbel-Copula, Gaussian-Copula and t-Copula functions, Based on the optimal joint distribution function, calculate the occurrence probability of dry-wet compound events between adjacent seasons under different scenarios.
[0011] More preferably, the calculation formulas for the occurrence probabilities of drought-to-wet transition, wet-to-drought transition, consecutive drought, and consecutive wet between adjacent seasons under the moderate scenario are respectively, [Formula] [Formula] [Formula] [Formula] , and the calculation formulas for the occurrence probabilities of drought-to-wet transition, wet-to-drought transition, consecutive drought, and consecutive wet between adjacent seasons under the severe scenario are respectively, [Formula] [Formula] [Formula] [Formula] , and Here, [Formula] denotes the joint distribution function of the SPI time series between adjacent seasons, X and Y respectively represent the SPI values of two events, X represents the precipitation status of the current season, and Y represents the precipitation status of the next season, [Formula] , [Formula] respectively denote the occurrence probabilities of the transition from drought to wet between adjacent seasons under moderate and severe scenarios, [Formula] , [Formula] respectively denote the occurrence probabilities of the transition from wet to drought between adjacent seasons under moderate and severe scenarios, [Formula] , [Formula] respectively denote the occurrence probabilities of consecutive drought between adjacent seasons under moderate and severe scenarios, [Formula] , [Formula] respectively denote the occurrence probabilities of consecutive wet conditions between adjacent seasons under moderate and severe scenarios.
[0012] More preferably, in step S20, based on the seasonal-scale time-series SPI index of precipitation data and monthly NDVI data, the Bayesian framework and the Copula function are combined to calculate the vegetation loss probability under the stress of dry-wet compound events between adjacent seasons under different scenarios.
[0013] More preferably, in step S20, the step of combining the Bayesian framework and the Copula function to calculate the probability of vegetation loss under stress from adjacent interseasonal dry-wet combined events under different scenarios is specifically as follows: The candidate marginal distributions are the normal distribution, the generalized extreme value distribution, and the Gumbel distribution, respectively. The candidate joint distribution functions are Clayton-Copula, Frank-Copula, Gumbel-Copula, Gaussian-Copula, and t-Copula. The optimal distribution for NDVI is selected using the KS test and the Akaike Information Criterion. The combined distribution function was selected using the root mean square error and AIC criteria. Next, based on the joint distribution function, we calculate the probability of vegetation loss under stress from adjacent interseasonal dry-wet combined events under different scenarios.
[0014] More preferably, in step S20, the formula for the probability of vegetation loss under stress from a drought-to-wet transition event between adjacent seasons under a moderate scenario is:
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
[0015] More preferably, in step S40, the risk of vegetation loss under the stress of a combined wet and dry event is calculated by raising the probability of occurrence and loss in step S20 to the power of the degree of exposure in step S30.
[0016] In a second aspect, the present invention provides a device for assessing vegetation loss risk under the stress of a combined drought-wet event, comprising a processor and a storage medium, the processor loading and executing instructions and data in the storage medium to realize a method for assessing vegetation loss risk under the stress of the rapid transition from drought to flood. [Effects of the Invention]
[0017] The beneficial effects of the second embodiment described above can be found in the explanation related to the first embodiment described above, and a further explanation is omitted here. [Brief explanation of the drawing]
[0018] [Figure 1] This flowchart outlines a method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood, as provided by this application. [Figure 2] This invention provides a spatial distribution pattern diagram of terrestrial vegetation exposure levels in China from 1982 to 2022. [Modes for carrying out the invention]
[0019] To clarify the purpose, technical solutions, and advantages of the embodiments of this application, the technical solutions of the embodiments of this application will be described clearly and completely below with reference to the drawings of the embodiments. Naturally, the embodiments described are only a selection of embodiments of this application, not all of them. Typically, the components of the embodiments of this application described and shown herein can be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of this application shown in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments that can be obtained by a person skilled in the art without creative work based on the embodiments of this application fall within the scope of protection of this application.
[0020] The following is an explanation of technical terms related to this application.
[0021] A disaster victim is an evaluation element in a disaster event, referring to an object that may be damaged or affected in a disaster event, and may include population, roads, plants and animals, buildings, etc., but in this application it refers to vegetation.
[0022] Vulnerability of a disaster-stricken object refers to the likelihood or extent to which it will suffer loss when affected by a disaster. In other words, it reflects the object's lack of sensitivity and adaptability to external disturbances (e.g., natural disasters, environmental changes, etc.) and is used to measure the object's "susceptibility to damage" or "degree of vulnerability" when a disaster occurs.
[0023] The degree of exposure of a damaged object refers to the extent or range to which an object may be threatened during a disaster, and is used to measure the distribution and quantitative scale of the object within the disaster's impact area.
[0024] This research has revealed that conventional disaster risk assessment methods have the following shortcomings.
[0025] (1) Limitations of independent event analysis
[0026] Traditional studies often focus on a single drought or flood event, neglecting the synergistic stress effects on vegetation from abrupt transitions from drought to flood (e.g., sudden heavy rainfall following a drought). For example, prolonged drought can reduce soil infiltration capacity, and subsequent heavy rainfall can cause more severe soil erosion and damage to vegetation root systems, but single-disaster models cannot quantify such nonlinear superimposed effects.
[0027] (2) Flaws in static vulnerability assessment
[0028] Existing studies often employ fixed vulnerability indicators (e.g., vegetation type, soil water retention capacity), but under conditions of rapid transition from drought to flood, vegetation's adaptive capacity changes dynamically with stress duration. For example, a short-term drought may only affect photosynthesis, but if accompanied by flooding, it can lead to root system oxygen deprivation and subsequent death, and conventional methods struggle to reflect such dynamic response processes.
[0029] (3) Uniformity of exposure degree representation
[0030] While research on the exposure levels of human systems (e.g., population, GDP) is relatively advanced, the exposure levels of natural systems (e.g., important ecological areas, vegetation function groups) are often neglected.
[0031] To address the above issues, this invention provides a method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood (a combined drought-wet event). This method quantitatively evaluates the risk of vegetation loss by comprehensively considering the probability of occurrence of the combined drought-wet event, the vulnerability (probability of loss) of vegetation, and the degree of exposure, and by utilizing data such as the Standardized Precipitation Index (SPI) and the Normalized Nucleotide Index (NDVI), and combining the Copula function and a Bayesian framework.
[0032] As shown in Figure 1, the method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood provided in this application includes steps S10 to S40, and the details are as follows.
[0033] S10. Monthly precipitation data and Normalized Density Index (NDVI) data for the target area are obtained and processed to obtain gridded seasonal time series of precipitation data and NDVI data.
[0034] In one embodiment, step S10 provided by the present invention specifically involves acquiring specified precipitation data and NDVI data, and processing the precipitation data and NDVI data into time series on a seasonal scale. In this embodiment, the seasonal division criteria are spring (March to May), summer (June to August), autumn (September to November), and winter (December to February of the following year). Next, the precipitation data and NDVI data are processed using spatial interpolation to obtain gridded precipitation data and NDVI data with high spatial resolution. Preferably, the spatial resolution here can be 0.5°.
[0035] S20 identifies inter-seasonal dry-wet composite events based on seasonal scale time series of precipitation data, categorizes them into moderate and severe scenarios, and calculates the probability of inter-seasonal dry-wet composite events occurring under both scenarios, as well as the probability of vegetation loss under the stress of inter-seasonal dry-wet composite events under both scenarios.
[0036] In one embodiment, the step of calculating the probability of occurrence of a combined dry / wet event between adjacent seasons under both scenarios provided by the present invention is specifically as follows:
[0037] (1) After obtaining the precipitation series in step S10, calculate the SPI index.
[0038] (2) Next, using the drought level classification criteria in Table 1, drought events and wet events are extracted by the SPI index, and different composite events are combined based on the season in which the drought and wet events occur. Spring-Summer transition from drought to wet, Spring-Summer transition from wet to drought, Spring-Summer continuous drought, Spring-Summer continuous wet, Summer-Autumn transition from drought to wet, Summer-Autumn transition from wet to drought, Summer-Autumn continuous drought, Summer-Autumn continuous wet, Autumn-Winter transition from drought to wet, Autumn-Winter transition from wet to drought, Autumn-Winter continuous drought, Autumn-Winter continuous wet, Winter-Spring transition from drought to wet, Winter-Spring transition from wet to drought, Winter-Spring continuous drought, Winter-Spring continuous wet.
[0039] [Table 1]
[0040] (3) Based on the levels of drought and wet events, examine the probability of occurrence of interseasonal combined dry-wet events under moderate and severe scenarios.
[0041] Specifically, the process for calculating the probability of a combined dry / wet event using a two-dimensional Copula function is as follows: First, the marginal distributions of the SPI indices of adjacent seasons are fitted using the Normal and Generalized Extreme Value (GEV) distributions. The Akaike Information Criterion (AIC) is used to select the optimal marginal distribution. Then, using the Root Mean Square Error (RMSE) and the AIC criterion, the optimal combined distribution function for the SPI time series between adjacent seasons is selected from among the Clayton-Copula, Frank-Copula, Gumbel-Copula, Gaussian-Copula, and t-Copula functions. Based on the selected optimal combined distribution function, the probability of a combined dry / wet event between adjacent seasons is calculated. The calculation formula is as follows: The formulas for calculating the probability of transitions between drought and wet seasons, transitions between wet and drought, consecutive droughts, and consecutive wet seasons under a moderate scenario are as follows:
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
[0042] In one embodiment, the step of calculating the probability of vegetation loss (vulnerability) under the stress of adjacent interseasonal wet-dry combined events in both scenarios provided by the present invention is specifically as follows:
[0043] Based on SPI time series and monthly NDVI time series, the probability of vegetation loss under stress from a combined dry-wet event is calculated by combining a Bayesian framework and the Copula function. The specific process is as follows: The candidate marginal distributions are the normal distribution, generalized extreme value distribution, and Gumbel distribution, respectively, and the candidate joint distribution functions are Clayton-Copula, Frank-Copula, Gumbel-Copula, Gaussian-Copula, and t-Copula. The optimal distribution of NDVI is selected using the KS test and the Akaike Information Criterion, and the joint distribution function is selected using the root mean square error and the AIC criterion. Based on the selected optimal joint distribution function, the probability of vegetation loss under stress from a combined dry-wet event is calculated. The calculation formula is as follows: The formula for the probability of vegetation loss under the stress of a drought-to-wet transition event between adjacent seasons under a moderate scenario is:
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
number
[0044] In S30, the vegetation system is treated as the affected entity, and the exposure of the vegetation system to a combined dry-wet event is expressed using the ground surface vegetation cover status. Specifically, the exposure of the vegetation system to a combined dry-wet event is calculated using a seasonally averaged normalized vegetation index based on the seasonal scale time series of NDVI data.
[0045] In step S40, based on the probability of occurrence and loss in step S20 and the degree of exposure in step S30, the risk of vegetation loss under the stress of a combined dry and wet event is calculated using the formula "Risk of Loss = Probability of occurrence of dry and wet combined event × Vulnerability (probability of loss) × degree of exposure". In other words, the probability of occurrence and loss in step S20 and the degree of exposure in step S30 are raised to the power of each other to calculate the risk of vegetation loss under the stress of a combined dry and wet event.
[0046] The present invention's method for evaluating vegetation loss risk under the stress of a rapid transition from drought to flood has the following effects: (1) By constructing a combined distribution of composite events and quantifying the dependence of composite events through composite event probability modeling based on the Copula function, it better fits the characteristics of actual disaster chains compared to the superposition of single disaster probabilities. (2) By prioritizing the exposure level of natural systems compared to exploring risk primarily based on human systems (population, GDP), it becomes easier to identify high-risk vegetation areas (e.g., wetland-grassland transition zones), improving spatial targeting for ecological management, and at the same time providing important reference value for preventing and controlling ecological risks under climate change.
[0047] The following describes in detail the method provided by this application for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood, using the nine major river basins of China as an example.
[0048] This specific embodiment provides a method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood. Based on the probability of a combined drought-wet event occurring (P), the probability of vegetation loss under the stress of a combined drought-wet event calculated in step S20, and the exposure level of the vegetation system calculated in step S30, the risk of vegetation loss under the stress of a combined drought-wet event is estimated by raising the probability of a combined drought-wet event occurring (P), the vulnerability of the vegetation (VuI), and the exposure level (Exp) to the power. The formula is as follows:
number
[0049] (1) Analysis of vegetation exposure
[0050] Based on monthly NDVI data from 1982 to 2022, the average NDVI value on a seasonal scale for each pixel is calculated to represent the vegetation's exposure to combined dry-wet events. A higher NDVI value indicates vigorous vegetation growth and high cover, meaning a higher exposure of the corresponding vegetation system to combined dry-wet events. Conversely, a lower NDVI value indicates poor vegetation growth and low cover, meaning a lower exposure of the corresponding vegetation system to combined dry-wet events. Figure 2 shows the spatial distribution pattern of vegetation exposure on a seasonal scale across the country from 1982 to 2022. Temporally, summer has the highest vegetation cover with an NDVI value of 0.45, followed by autumn with relatively good vegetation growth and an NDVI value of 0.36. Due to the influence of other factors such as temperature and sunlight, vegetation cover is low in spring and winter, with NDVI values of 0.31 and 0.25, respectively. Spatially, the NDVI value gradually increases from northwest to southeast, with the NDVI value concentrated in the northwest at 0-0.4 and the NDVI value mainly distributed in the southeast at 0.4-1.0. As is clear from Table 2, the Pearl River basin and the southeastern river basins have good vegetation growth throughout the seasons, while the Yellow River basin and inland river basins have poor growth. Therefore, the spatial differences in vegetation exposure among the various river basins in China are relatively significant.
[0051] [Table 2]
[0052] (2) Mean risk of vegetation loss under the stress of combined dry and wet events
[0053] The average risk of vegetation loss under the stress of a combined dry-wet event is calculated according to the formula "Risk = Risk of disaster factor × Vulnerability × Exposure". The probability of occurrence of the combined dry-wet event calculated in step S20 is used as the risk of disaster factor, the probability of vegetation loss under the stress of the combined dry-wet event is used as vulnerability, and the vegetation exposure calculated in step S30 is used.
[0054] (2.1) Mean risk of summer vegetation loss under the stress of combined spring-summer wet-dry events
[0055] Under the moderate scenario, as is evident from Table 3, spatially, the Songliao River basin had the highest mean risk of summer vegetation loss under the stress of a spring-summer combined dry-wet event (3.35), followed by the southeastern river basins (3.3), the Yangtze River basin (3.08), the Pearl River basin (2.95), the Huai River basin (2.71), the Hai River basin (2.67), the Yellow River basin (2.15), and the southwestern river basins (2.15), with the inland river basins having the lowest mean risk (0.79). Analyzing different events, the mean risk for continuous drought events was 0.83%, the mean risk for continuous wet events was 0.52%, and the mean risks for drought-to-wet transition and wet-to-drought transition events were 0.41% and 0.32%, respectively. This indicates that the mean risks for continuous drought and continuous wet events are higher than those for dry-wet transition events. Observing the distribution of high-risk areas for different drought-to-wet combined events under moderate scenarios, high-risk areas for drought-to-wet transition events are located in the northern part of the Songliao River basin, high-risk areas for wet-to-drought transition events are located in the southern part of the southeastern river basins, and high-risk areas for continuous drought and continuous wet events are essentially the same, both concentrated in the southeastern river basins and the northwestern border region of the inland river basins of China. Notably, the distribution of high-risk areas for drought-to-wet transition and continuous drought is almost inverse.
[0056] Under severe scenarios, as is evident from Table 3, spatially, the mean risk of summer vegetation loss under the stress of spring-summer combined dry-wet events was highest in the southeastern river basins (0.71), followed by the Songliao River basin (0.7), the Yangtze River basin (0.68), the Hai River basin (0.58), the Pearl River basin (0.56), the Huai River basin (0.54), the Yellow River basin (0.5), and the southwestern river basins (0.48). The inland river basins had the lowest mean risk (0.18). Analyzing different events, the mean risk for continuous drought events was 0.21%, the mean risk for continuous wet events was 0.11%, and the mean risks for drought-to-wet transition and wet-to-drought transition events were 0.08% and 0.05%, respectively. This indicates that the mean risks for continuous drought and continuous wet events are higher than those for dry-wet transition events. Observing the distribution of high-risk areas for different drought-to-wet combined events under severe scenarios, no clear high-risk areas were observed for drought-to-wet transition and wet-to-drought transition events. High-risk areas for continuous drought and continuous wet events were essentially the same, both concentrated in the western parts of China's Yangtze and Yellow River basins.
[0057] [Table 3]
[0058] (2.2) Mean risk of autumn vegetation loss under the stress of combined spring-summer wet-dry events
[0059] Under the moderate scenario, as is evident from Table 4, spatially, the mean risk of autumn vegetation loss under the stress of spring-summer combined dry-wet events was highest in the southeastern river basins (3.24), followed by the Pearl River basin (3.14), Yangtze River basin (2.62), Huai River basin (1.99), southwestern river basins (1.96), Songliao River basin (1.92), Hai River basin (1.69), and Yellow River basin (1.48), with the lowest mean risk in the inland river basins (0.55). Analyzing different events, the mean risk for continuous drought events was 0.63%, the mean risk for continuous wet events was 0.39%, and the mean risks for drought-to-wet transition and wet-to-drought transition events were 0.29% and 0.28%, respectively. This indicates that the mean risks for continuous drought and continuous wet events are higher than those for dry-wet transition events. Observing the distribution of high-risk areas for different drought-to-wet combined events under moderate scenarios, high-risk areas for drought-to-wet transition events are located in the northern part of the Songliao River basin; high-risk areas for wet-to-drought transition events are located in the southern part of the Southeast River basins and the lower reaches of the Pearl River basin; high-risk areas for continuous drought events are concentrated in the northwestern border regions of China's southeastern and inland river basins; and high-risk areas for continuous wet events are concentrated in the upper and lower reaches of the Yangtze River basin, the lower reaches of the Southwest River basins, and the northern part of the Southeast River basins.
[0060] Under severe scenarios, as is evident from Table 4, spatially, the mean risk of summer vegetation loss under the stress of spring-summer combined dry-wet events was highest in the southeastern river basins (0.73), followed by the Pearl River basin (0.6), Yangtze River basin (0.58), southwestern river basins (0.43), Songliao River basin (0.42), Huai River basin (0.41), Hai River basin (0.38), and Yellow River basin (0.34), with the lowest mean risk in the inland river basins (0.13). Analyzing different events, the mean risk for continuous drought events was 0.16%, the mean risk for continuous wet events was 0.08%, and both drought-to-wet and wet-to-drought transition events were 0.05%. This indicates that the mean risks for continuous drought and continuous wet events are higher than those for dry-wet transition events. Observing the distribution of high-risk areas for different drought-wet combined events under severe scenarios, no clear high-risk areas were found for drought-to-wet transition, wet-to-drought transition, and continuous wet events. High-risk areas for continuous drought events were distributed in the western parts of the Yangtze and Yellow River basins in China.
[0061] [Table 4]
[0062] Comparing the mean risk of summer vegetation loss with the mean risk of autumn vegetation loss under the stress of a combined spring-summer dry-wet event, it can be seen that the mean risk of summer vegetation loss is higher than the mean risk of autumn vegetation loss in both moderate and severe scenarios. Looking at different regions, the inland river basins, Songliao River basin, Yellow River basin, Hai River basin, and Huai River basin all satisfy the characteristic that the average risk of summer vegetation loss is higher than the average risk of autumn vegetation loss. However, in the Yangtze River basin during severe wet-to-drought transitions, the Pearl River basin during moderate wet-to-drought transitions and continuous droughts, the Pearl River basin during severe continuous droughts, the Southwest River basins during moderate and severe wet-to-drought transitions, and the Southeast River basins during moderate and severe wet-to-drought transitions and continuous wetness, the average risk of autumn vegetation loss is higher than the average risk of summer vegetation loss. This indicates that all basins where the average risk of autumn vegetation loss is higher than the average risk of summer vegetation loss are located in the south and are all concentrated in wet-to-drought transition events.
[0063] (2.3) Mean risk of autumn vegetation loss under the stress of a combined summer-autumn wet-dry event.
[0064] Under the moderate scenario, as is evident from Table 5, spatially, the Pearl River basin had the highest average risk of autumn vegetation loss under the stress of summer-autumn combined dry-wet events (3,2), followed by the southeastern river basins (3,13), the Yangtze River basin (2,66), the Huai River basin (2,08), the southwestern river basins (2,04), the Songliao River basin (1,90), the Hai River basin (1,69), and the Yellow River basin (1,49), with the inland river basins having the lowest average risk (0.55). Analyzing different events, the average risk for continuous drought events was 0.61%, the average risk for continuous wet events was 0.49%, and the average risks for drought-to-wet transition and wet-to-drought transition events were 0.28% and 0.22%, respectively. This indicates that the average risks for continuous drought and continuous wet events are higher than those for dry-wet transition events. Observing the distribution of high-risk areas for different drought-to-wet combined events under moderate scenarios, high-risk areas for drought-to-wet transition events are located in the lower reaches of the Yangtze River basin, the lower reaches of the Pearl River basin, and the southern parts of the Southeast River basins. No clear high-risk areas are observed for wet-to-drought transition events. For continuous drought and continuous wet events, the high-risk areas in the Yangtze River basin, Southeast River basin, Southwest River basin, and Pearl River basin are essentially the same, while the high-risk areas in the Songliao River basin, Haihe River basin, and Huaihe River basin are almost the opposite.
[0065] Under severe scenarios, as is evident from Table 5, spatially, the Pearl River basin had the highest average risk of autumn vegetation loss under the stress of summer-autumn combined dry-wet events (0.64), followed by the southeastern river basins (0.61), the Yangtze River basin (0.60), the southwestern river basins (0.47), the Huai River basin (0.43), the Songliao River basin (0.43), the Hai River basin (0.35), and the Yellow River basin (0.35), with the inland river basins having the lowest average risk (0.12). Analyzing different events, the average risk for continuous drought events was 0.16%, the average risk for continuous wet events was 0.11%, and the drought-to-wet transition and wet-to-drought transition events were 0.05% and 0.03%, respectively. This indicates that the average risks for continuous drought and continuous wet events are higher than those for dry-wet transition events. Observing the distribution of high-risk areas for different drought-wet combined events under severe scenarios, no clear high-risk areas were found for drought-to-wet transition, wet-to-drought transition, and continuous wet events. High-risk areas for continuous drought events were distributed in the western parts of the Yangtze and Yellow River basins in China.
[0066] [Table 5]
[0067] This specific example evaluated the risk of vegetation loss under the stress of a combined dry-wet event, based on a definition of vegetation loss risk. The following conclusions were reached.
[0068] (1) High-risk areas for mean risk of summer vegetation loss under moderate spring-summer drought-to-humidity transition events are located in the northern Songliao River basin, high-risk areas for humid-to-drought transition events are distributed in the southern parts of the southeastern river basins, and high-risk areas for continuous drought and continuous humidity events basically coincide and are concentrated in the southeastern river basins and the northwestern border region of the inland river basins of China. Notably, the high-risk areas for drought-to-humidity transition and continuous drought show almost opposite distributions. Under severe scenarios, high-risk areas for continuous drought and continuous humidity events basically coincide and are concentrated in the western parts of the Yangtze River and Yellow River basins of China.
[0069] (2) Under the stress of a moderate spring-summer drought-to-wet transition event, high-risk areas for the mean risk of autumn vegetation loss are located in the northern Songliao River basin, high-risk areas for a wet-to-drought transition event are located in the southern Southeast River basins and the lower Pearl River basin, high-risk areas for a continuous drought event are concentrated in the southeastern river basins and the northwestern border region of the inland river basins of China, and high-risk areas for a continuous wet event are concentrated in the upper and lower reaches of the Yangtze River basin, the lower reaches of the Southwest River basins, and the northern part of the Southeast River basins. Under the severe scenario, high-risk areas for a continuous drought event are located in the western parts of the Yangtze River basin and the Yellow River basin of China. Comparing the mean risk of summer vegetation loss and the mean risk of autumn vegetation loss under the stress of a combined spring-summer dry-wet event, it can be seen that in both the moderate and severe scenarios, the mean risk of summer vegetation loss is higher than the mean risk of autumn vegetation loss.
[0070] (3) High-risk areas for mean risk of autumn vegetation loss under moderate summer-autumn drought-to-wet transition events are located in the lower Yangtze River basin, the lower Pearl River basin, and the southern parts of the Southeast River basins. No clear high-risk areas are observed for wet-to-drought transition events. High-risk areas for continuous drought and continuous wet events are basically the same in the Yangtze River basin, Southeast River basin, Southwest River basin, and Pearl River basin, while the high-risk areas in the Songliao River basin, Haihe River basin, and Huaihe River basin are almost the opposite. Under severe scenarios, high-risk areas for continuous drought events are distributed in the western parts of the Yangtze River and Yellow River basins in China.
[0071] As will be readily apparent to those skilled in the art, the foregoing are merely preferred embodiments of the Application and do not limit it. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the Application shall also be included within the scope of protection of the Application.
Claims
1. A method for assessing the risk of vegetation loss under the stress of a rapid transition from drought to flood, S10, the steps of obtaining monthly precipitation data and NDVI data for the target area, processing them to obtain seasonal scale time series of gridded precipitation data and NDVI data, S20, based on a time series of precipitation data on a seasonal scale, identify inter-seasonal dry-wet composite events, categorize them into moderate and severe scenarios, then use the Copula function to calculate the probability of inter-seasonal dry-wet composite events occurring under different scenarios, and simultaneously combine the Bayesian framework and the Copula function to calculate the probability of vegetation loss under stress for inter-seasonal dry-wet composite events under different scenarios. S30, a step of calculating the exposure of the vegetation system to a combined dry and wet event using a seasonally averaged normalized vegetation index based on a seasonal time series of NDVI data, A method for assessing vegetation loss risk under the stress of a rapid transition from drought to flood, comprising the step of calculating the risk of vegetation loss under the stress of a combined dry-wet event based on the probability of occurrence and the probability of loss calculated in step S20 and the degree of exposure calculated in step S30.
2. In step S20, the steps of identifying inter-seasonal dry-wet combined events and classifying them into moderate and severe scenarios based on seasonal time series of precipitation data are specifically as follows: Based on the seasonal time series of precipitation data, the SPI index is calculated. Based on the drought level classification criteria, drought events and wet events are extracted using the SPI index and classified into moderate and severe scenarios. A method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood, as described in claim 1, characterized in that, based on the seasons in which drought and wet events occur, a combined dry-wet event between adjacent seasons is identified, and the combined event includes four of the following: a transition from drought to wet from spring to summer, a transition from wet to drought from spring to summer, a continuous drought from spring to summer, a continuous wet from spring to summer, a transition from drought to wet from summer to autumn, a transition from wet to drought from summer to autumn, a continuous drought from summer to autumn, a continuous wet from summer to autumn, a transition from drought to wet from autumn to winter, a transition from wet to drought from autumn to winter, a continuous wet from autumn to winter, a transition from drought to wet from winter to spring, a transition from wet to drought from winter to spring, a continuous drought from winter to spring, and a continuous wet from winter to spring.
3. The method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood, as described in claim 2, characterized in that the drought level classification criteria are: -2 < SPI ≤ -1.5 is severe drought, -1.5 < SPI ≤ -1 is moderate drought, 1 < SPI ≤ 1.5 is moderate flood, and 1.5 < SPI ≤ 2 is severe flood.
4. In step S20, the step of calculating the probability of a combined dry-wet event occurring between adjacent seasons under different scenarios using the Copula function is specifically as follows: The marginal distributions of SPI indices for adjacent seasons are fitted using the normal distribution and the generalized extreme value distribution, and the optimal marginal distribution is selected using the Akaike Information Criterion. Using the root mean square error and AIC criteria, the optimal combined distribution function for adjacent seasons of SPI time series is selected from among the Clayton-Copula, Frank-Copula, Gumbel-Copula, Gaussian-Copula, and t-Copula functions. A method for assessing the risk of vegetation loss under the stress of a rapid transition from drought to flood, according to claim 1, characterized by calculating the probability of occurrence of a combined dry-wet event between adjacent seasons under different scenarios, based on an optimal combined distribution function.
5. The formulas for calculating the probability of transitions between drought and wet seasons, transitions between wet and drought, consecutive droughts, and consecutive wet seasons under a moderate scenario are as follows: [Math 1] [Math 2] [Math 3] [Math 4] And, The formulas for calculating the probability of transitions between drought and wet conditions, wet and drought conditions, consecutive droughts, and consecutive wet conditions between adjacent seasons under severe scenarios are as follows: [Math 5] [Math 6] [Number 7] [Number 8] And, Here, [Number 9] The graph shows the combined distribution function of SPI time series between adjacent seasons, where X and Y represent the SPI values of two events, where X represents the precipitation conditions for the current season, and where Y represents the precipitation conditions for the next season. [Number 10] 、 [Math 11] These represent the probability of drought-to-wet transition occurring between adjacent seasons under moderate and severe scenarios, respectively. [Math 12] 、 [Number 13] These represent the probability of a transition from wet to drought between adjacent seasons under moderate and severe scenarios, respectively. [Number 14] 、 [Equation 15] These represent the probability of consecutive droughts between adjacent seasons under moderate and severe scenarios, respectively. [Number 16] 、 [Number 17] A method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood, according to claim 1, characterized in that the probabilities of consecutive wet periods occurring between adjacent seasons under moderate and severe scenarios, respectively.
6. A method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood, according to claim 1, characterized in that, in step S20, the probability of vegetation loss under the stress of a combined dry-wet event between adjacent seasons is calculated by combining a Bayesian framework and a Copula function based on the SPI index of seasonal scale time series precipitation data and monthly NDVI data.
7. In step S20, the steps of combining the Bayesian framework and the Copula function to calculate the probability of vegetation loss under stress from adjacent interseasonal dry-wet combined events under different scenarios are specifically as follows: The candidate marginal distributions are the normal distribution, the generalized extreme value distribution, and the Gumbel distribution, respectively. The candidate joint distribution functions are Clayton-Copula, Frank-Copula, Gumbel-Copula, Gaussian-Copula, and t-Copula. The optimal distribution of NDVI is selected using the K-S test and the Akaike Information Criterion. The combined distribution function was selected using the root mean square error and AIC criteria. A method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood, according to claim 1, further comprising calculating the probability of vegetation loss under the stress of a combined dry-wet event between adjacent seasons under different scenarios, based on a combined distribution function.
8. In step S20, the formula for the probability of vegetation loss under the stress of an adjacent interseasonal drought-to-wet transition event under a moderate scenario is: [Number 18] [Number 19] [Number 20] And, The formula for the probability of vegetation loss under stress from a wet-to-drought transition event between adjacent seasons under a moderate scenario is: [Math 21] [Number 22] [Number 23] And, The formula for the probability of vegetation loss under stress from consecutive interseasonal drought events in a moderate scenario is: [Number 24] [Number 25] [Number 26] And, The formula for the probability of vegetation loss under stress from consecutive interseasonal wet events in a moderate scenario is: [Number 27] [Number 28] [Number 29] And, The formula for the probability of vegetation loss under the stress of a transition event from drought to wet conditions between adjacent seasons under a severe scenario is: [Number 30] [Number 31] [Number 32] And, The formula for the probability of vegetation loss under the stress of a transition event from wet to drought between adjacent seasons under a severe scenario is: [Number 33] [Number 34] [Number 35] And, The formula for the probability of vegetation loss under stress from consecutive interseasonal drought events in a severe scenario is: [Number 36] [Number 37] [Number 38] And, The formula for the probability of vegetation loss under stress from consecutive interseasonal wet events in a severe scenario is: [Number 39] [Number 40] [Number 41] And, Here, X represents the NDVI time series, and Y and Z represent the SPI time series and the NDVI time series, respectively. [Number 42] The function shown represents the marginal distribution function of the NDVI data series, where x is NDVI < NDVI. 40th , in other words, it shows a vegetation loss scenario, [Number 43] and [Number 44] This shows the marginal distribution functions of the SPI time series for two seasons. [Number 45] This shows the combined distribution function of the SPI time series for two seasons. [Number 46] This shows the combined distribution function of the SPI and NDVI time series for two seasons. [Number 47] 、 [Number 48] These represent the probability of vegetation loss under the stress of interseasonal drought-to-wet transitions under moderate and severe scenarios, respectively. [Number 49] 、 [Number 50] These represent the probability of vegetation loss under the stress of interseasonal wet-to-drought transitions under moderate and severe scenarios, respectively. [Number 51] 、 [Number 52] These represent the probability of vegetation loss under the stress of consecutive interseasonal droughts under moderate and severe scenarios, respectively. [Number 53] 、 [Number 54] A method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood, according to claim 1, characterized in that the probabilities of vegetation loss under the stress of continuous wetness between adjacent seasons under moderate and severe scenarios, respectively.
9. A method for evaluating the risk of vegetation loss under the stress of a rapid transition from drought to flood, according to claim 1, characterized in that in step S40, the risk of vegetation loss under the stress of a combined dry-wet event is calculated by raising the probability of occurrence and the probability of loss in step S20 to the power of the degree of exposure in step S30.
10. A vegetation loss risk assessment device under the stress of a combined dry-wet event, comprising a processor and a storage medium, wherein the processor loads and executes instructions and data in the storage medium to realize a method for assessing the risk of vegetation loss under the stress of a rapid transition from drought to flood as described in any one of claims 1 to 9.
Citation Information
Patent Citations
Drought index monitoring method and system based on deep learning
CN112668705A
Multivariable rasterizing satellite remote sensing comprehensive drought risk assessment method
CN115099453A
Quantitative identification method for vegetation loss probability under stress of composite climate event
CN118708894A
Forest drought monitoring method, equipment and medium
CN119167207A
Drought risk identification early warning method based on multi-source satellite fusion NDVI and machine learning
CN120031207A
