Vegetation drought index evaluation method and system based on water vapor flux divergence

Abstract

The present application belongs to the technical field of agricultural meteorological disasters, and particularly relates to a vegetation drought disaster index evaluation method and system based on water vapor flux divergence. The evaluation method comprises the following steps: obtaining historical data of water vapor flux divergence of a target region, processing the historical data to obtain source historical data, calculating the occurrence probability of water vapor flux divergence according to the source historical data, obtaining the water vapor flux divergence index from the occurrence probability, obtaining a satellite remote sensing vegetation change index according to a satellite remote sensing vegetation index, calculating the significance region of the water vapor flux divergence index and the satellite remote sensing vegetation change index, and dividing the vegetation drought disaster level of the significance region according to the water vapor flux divergence index. The present application solves the problems that the traditional drought index does not consider all the factors affecting drought disasters, resulting in low accuracy and mutual interference among multiple vegetation drought disaster factors, and that data processing and analysis are difficult, and overcomes the defect that the complex surface does not have drought disaster prediction capability.

Description

Technical Field

[0001] The present invention belongs to the technical field of agricultural meteorological disasters, and in particular relates to a vegetation drought index assessment method and system based on water vapor flux divergence. Background Art

[0002] Drought occurs when terrestrial production systems are negatively impacted by an imbalance in water revenue and expenditure. Unlike meteorological and climatic droughts, vegetation droughts occur when there is no rain or very little rain, causing a drop in soil moisture. Plants suffer because they cannot get the water they need. In addition to being affected by the amount of water in the environment, plant droughts are also related to their own drought resistance. In existing technologies, common prediction methods primarily assess drought from a water-scarce perspective based on precipitation and evapotranspiration, such as using a drought index based on factors like precipitation, evapotranspiration, and water vapor content. Other methods assess drought responses to remotely sensed vegetation. These methods often suffer from the following drawbacks: poor anti-interference capabilities due to complex surfaces, low accuracy, and a lack of predictive power. Currently, there is no method that comprehensively covers the factors affecting vegetation droughts and can effectively, conveniently, and accurately assess and predict vegetation droughts. Summary of the Invention

[0003] In view of the above analysis and in response to the deficiencies in the existing technology, the present invention aims to provide a vegetation drought index assessment method and system based on water vapor flux divergence, which can at least solve one of the following technical problems: (1) The traditional drought index does not fully consider the factors affecting drought in complex surface vegetation, resulting in low accuracy; (2) In the traditional drought index, multiple vegetation drought influencing factors interfere with each other, making data processing and analysis difficult; (3) Vegetation drought monitoring using satellite remote sensing technology is easily interfered with by insect pests, human factors, etc.

[0004] The purpose of the present invention is mainly achieved through the following technical solutions:

[0005] The present invention provides a vegetation drought index assessment method based on water vapor flux divergence, comprising the following steps:

[0006] Obtain historical data of water vapor flux divergence in the target area, and obtain source historical data based on the historical data processing; the historical data includes water vapor flux divergence data of the target area obtained by querying meteorological data; the source historical data includes water vapor flux divergence data of the required time unit;

[0007] The probability of occurrence of water vapor flux divergence at a specified time is calculated based on historical source data;

[0008] Calculate the water vapor flux divergence index at a specified time according to the occurrence probability of water vapor flux divergence at a specified time;

[0009] Satellite remote sensing vegetation index of the target region is acquired, satellite remote sensing vegetation variation index is obtained according to the satellite remote sensing vegetation index, and the significance region of the water vapor flux divergence index and the satellite remote sensing vegetation variation index of the target region at the specified time is obtained by linear regression analysis according to the obtained water vapor flux divergence index at the specified time and the satellite remote sensing vegetation variation index.

[0010] The vegetation drought grade of the significance region is divided according to the water vapor flux divergence index at the specified time.

[0011] Preferably, the calculation method of the occurrence probability p of the water vapor flux divergence at the specified time is as follows:

[0012]

[0013] wherein VIMDy,m is the water vapor flux divergence of the specified yth year and mth month; m is the number of 1-12 months; p(VIMDy,m) is the probability of the water vapor flux divergence of the specified yth year and mth month; nm is the sample number of the water vapor flux divergence of the mth month of different years in the same period of multiple years +1; iy,m is defined as follows: the water vapor flux divergence of the specified yth year and mth month and the data set of the water vapor flux divergence of the mth month of different years in the same period of multiple years are arranged in ascending order according to the numerical value, and the sequence numbers of 1, 2, …, nm are sequentially assigned, and iy,m is the sequence number of the water vapor flux divergence of the yth year and mth month in the arrangement.

[0014] Preferably, the calculation method of the water vapor flux divergence index at the specified time is as follows:

[0015] The occurrence probability p of the water vapor flux divergence VIMD is divided into two cases of ≤0.5 and >0.5, and the water vapor flux divergence is normalized according to the normal distribution to obtain the water vapor flux divergence normal distribution normalization index t, t satisfies:

[0016]

[0017] The water vapor flux divergence index SVIMDI satisfies:

[0018]

[0019] Preferably, the method for acquiring the satellite remote sensing vegetation variation index VA of the target region is as follows:

[0020] y is the year, m is the number of 1-12 months, VAy,m is the remote sensing vegetation variation index of the specified yth year and mth month, VIy,m is the vegetation index of the specified yth year and mth month, and VIave,m is the average value of the vegetation index of the mth month in the years.

[0021] Preferably, the method for determining the significant areas of the water vapor flux divergence index and the satellite remote sensing vegetation change index is: the water vapor flux divergence index and the satellite remote sensing vegetation change index are screened according to the F overall significance test, and the areas with a significance level less than 0.05 are determined as significant areas.

[0022] Preferably, before the F overall significance test, the target area can be divided into several minimum units from which the water vapor flux divergence index and / or the satellite remote sensing vegetation change index can be obtained, and the minimum units are subjected to the F overall significance test to screen the areas with a significance level less than 0.05 as significant areas.

[0023] Preferably, a linear regression model is constructed using the water vapor flux divergence index of the significant area as a prediction factor and the satellite remote sensing vegetation change index of the significant area as a prediction quantity.

[0024] Preferably, the classification standard for the vegetation drought disaster level is: SVIMDI>1.0, then the target area or the smallest unit of the target area is determined to be a wet area; -1.0<SVIMDI≤1.0, then the target area or the smallest unit of the target area is determined to be a normal area; SVIMDI≤-1.0, then the target area or the smallest unit of the target area is determined to be a drought area.

[0025] Preferably, the original historical data includes at least 15 years of water vapor flux divergence data.

[0026] On the other hand, the present invention discloses a vegetation drought index assessment system based on water vapor flux divergence, which is used to implement the above-mentioned assessment method and includes the following modules:

[0027] Data acquisition and processing module:

[0028] Used to obtain historical data of water vapor flux divergence in the target area, and to obtain source historical data based on historical data processing; the historical data includes water vapor flux divergence data of the target area obtained through meteorological data query; the source historical data includes water vapor flux divergence data of the required time unit;

[0029] First calculation module:

[0030] The probability of occurrence of water vapor flux divergence at a specified time is calculated based on historical source data;

[0031] Second calculation module:

[0032] Calculate the water vapor flux divergence index at a specified time according to the occurrence probability of water vapor flux divergence at a specified time;

[0033] The third calculation module:

[0034] Obtain the satellite remote sensing vegetation index of the target area, obtain the satellite remote sensing vegetation change index based on the satellite remote sensing vegetation index, and perform linear regression analysis based on the obtained water vapor flux divergence index and satellite remote sensing vegetation change index at the specified time to obtain the significant area of ​​the water vapor flux divergence index and satellite remote sensing vegetation change index at the specified time in the target area;

[0035] Drought level determination module:

[0036] The vegetation drought level of significant areas is divided according to the water vapor flux divergence index at a specified time.

[0037] Compared with the prior art, the present invention can achieve at least one of the following technical effects:

[0038] (1) The present invention uses water vapor flux divergence as the analysis variable. Water vapor flux divergence regards the region as a whole to characterize water vapor input and output, and can reflect the comprehensive influence of factors such as precipitation and evapotranspiration. Therefore, compared with the common drought index that simply considers precipitation, evaporation, and soil moisture, water vapor flux divergence can comprehensively consider various factors affecting vegetation drought, and it is easier to obtain the most realistic assessment results on complex surfaces.

[0039] (2) The present invention avoids the irrelevant factors of the water vapor flux divergence and the actual growth status of vegetation in the water vapor flux divergence assessment by analyzing the significance of the water vapor flux divergence index and the satellite remote sensing vegetation change index in the target area, thereby avoiding the one-sided description of drought from the perspective of atmospheric water shortage. At the same time, it takes into account the different tolerance levels of different plants to drought and the essential differences between plant drought and environmental drought, thereby greatly improving the accuracy of plant drought assessment.

[0040] (3) Compared with the existing method of using satellite remote sensing to monitor vegetation growth and then assess drought disasters, the present invention avoids the interference of insect pests, temperature, human factors, etc. on vegetation drought disaster assessment and has higher specificity and accuracy.

[0041] (4) Compared with the vegetation growth status assessment method such as satellite remote sensing, the present invention adopts the characteristic of water vapor flux divergence index's pre-emptive effect on vegetation drought. Changes in water vapor flux divergence first lead to changes in atmospheric moisture, which require feedback adjustment of moisture by vegetation before affecting vegetation growth. Therefore, the present invention uses water vapor flux divergence as a method for assessing vegetation drought, which can predict the vegetation growth status in a region within a certain period of time in the future.

[0042] (5) In the present invention, when calculating the probability of occurrence of water vapor flux divergence at a specified time, the water vapor flux divergence of the multi-year period in the relevant source historical data and the water vapor flux divergence of the specified time are sorted from small to large, covering all sample intervals of the source historical data. Compared with the partial sampling of historical data by the sampling method commonly used in the prior art, the present invention has a higher accuracy rate. At the same time, the present invention only mathematically sorts the sample intervals, does not require more complex calculations, and can be easily implemented with the assistance of a computer, thereby increasing the accuracy without excessively increasing the workload. At the same time, compared with another common method in the prior art that determines the specific distribution of probability through mathematical full analysis, the data processing workload is greatly simplified. The water vapor flux divergence at a specified time and the water vapor flux divergence of the same month in the multi-year period are arranged from small to large and compared through a formula to obtain the probability of water vapor flux divergence at a specified time.

[0043] Other features and advantages of the present invention will be described in the following description, and part of them may become obvious from the description or be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structure indicated in the written description and the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a flow chart of a method for evaluating vegetation drought index in an embodiment of the present invention;

[0045] Figure 2 Schematic diagram of the significant correlation region between SVIMDI and VA in an embodiment of the present invention;

[0046] Figure 3 Schematic diagram of the changes in drought levels in Kizilsu Kirgiz Autonomous Prefecture over the years in an embodiment of the present invention;

[0047] Figure 4 Schematic diagram of the significant correlation region between SPI and VA in the comparative example of the present invention;

[0048] Figure 5 This is a schematic diagram of the changes in drought levels in Kizilsu Kirgiz Autonomous Prefecture over the years in the comparative example of the present invention. DETAILED DESCRIPTION

[0049] The following is a further detailed description of the vegetation drought index assessment method and system based on water vapor flux divergence in conjunction with specific embodiments and comparative examples. These embodiments are only for comparison and explanation purposes, and the present invention is not limited to these embodiments.

[0050] The inventors discovered that, in the prior art, drought indices primarily assess drought from a water shortage perspective based on precipitation and evapotranspiration. These single indicators, such as precipitation, evapotranspiration, and water vapor content, fail to fully consider all factors influencing drought and ignore the actual impact of the plant's inherent drought tolerance on vegetation drought assessment. Consequently, the drought indices used in the prior art affect prediction or assessment accuracy. Furthermore, when combining factors such as precipitation, evapotranspiration, and water vapor content for analysis, these variables interact with each other, creating significant data processing difficulties and making it difficult to efficiently and effectively predict vegetation drought. Furthermore, traditional field survey methods for vegetation require significant manpower and material resources, making them difficult to meet the needs of real-time monitoring. While satellite remote sensing can monitor surface vegetation and environmental changes, it can only reveal the state of vegetation growth and cannot explain the causes of changes in vegetation growth (for example, pests and diseases and human factors can also affect vegetation growth). Furthermore, satellite remote sensing is based on monitoring the current state of vegetation growth and cannot predict trends in vegetation growth or its impact from drought.

[0051] Based on these findings, the present invention proposes a vegetation drought index assessment method and system based on water vapor flux divergence. Water vapor flux divergence characterizes the total convergence and divergence of water vapor in a given region. When water vapor is transported from a source to a region, if the horizontal input of water vapor exceeds the output, this manifests as net horizontal convergence of water vapor; if the horizontal input exceeds the output, this manifests as net horizontal divergence of water vapor. Because any indicator that can cause water vapor flow can affect water vapor flux divergence, which is the result of the combined effects of factors such as precipitation and evapotranspiration, calculating water vapor flux divergence over a given period of time in a given region can yield a plant drought assessment index that comprehensively considers these factors.

[0052] The present invention proposes a vegetation drought index assessment method and system based on water vapor flux divergence, such as Figure 1 Shown, including:

[0053] Step 1: Obtain historical data on water vapor flux divergence in the target area, and obtain source historical data based on historical data processing; historical data include water vapor flux divergence data of the target area obtained by querying the recorded meteorological data; source historical data include water vapor flux divergence data of the target area in the required time unit converted from the water vapor flux divergence data of the target area; specifically, historical data on water vapor flux divergence in the target area can be obtained through meteorological data query; in order to improve the accuracy and reliability of the evaluation results, historical data of multiple years can be obtained. For example, the historical data can include data of the target area for more than 15 consecutive years, and the obtained historical data can be converted into source historical data that can be further processed, such as monthly average data: historical data can be monthly data, quarterly data, or short-term data with a time period of days or even hours; monthly data and quarterly data can be directly used as source historical data; short-term data with a time period of days or even hours can be converted into monthly average data or quarterly data by using weighted average and other methods as needed.

[0054] Compared with the common drought index in existing technologies, the water vapor flux divergence represents the regional water vapor input and output, and uses a holistic approach to consider various factors affecting vegetation drought. This reduces the difficulty of combining variables such as precipitation and solar evaporation, which are mutually influential and selected in existing technologies, reduces the amount of data processing, and improves processing efficiency.

[0055] Step 2: Calculate the probability of water vapor flux divergence at a specified time based on historical source data;

[0056] Among them, the specified time is the time point at which the drought status of vegetation in the target area is to be evaluated. The specified time can be determined as a certain month of a certain year or a certain quarter of a certain year. For example, the mth month of the yth year to be evaluated in the historical data is selected as the specified time, and the water vapor flux divergence at the specified time and the multi-year water vapor flux divergence of the same month of the specified time are queried in the source historical data; the multi-year period refers to the time point in the historical year that is the same as the specified time and month; specifically, if the drought status of vegetation in the target area in May 2019 is to be evaluated, the dataset of the source historical data of water vapor flux divergence from May 1982 to 2018 can be selected as the multi-year data of water vapor flux divergence data in May 2019; specifically, if the drought status of vegetation in the target area in the first quarter of 2019 is to be evaluated, the dataset of the source historical data of water vapor flux divergence from the first quarter of 1982 to 2018 can be selected as the multi-year data of water vapor flux divergence data in the first quarter of 2019.

[0057] Specifically, based on the source historical data of water vapor flux divergence, the water vapor flux divergence at a specified time and the corresponding multi-year water vapor flux divergence are extracted, and the occurrence probability of the water vapor flux divergence at a specified time is calculated based on the two water vapor flux divergences.

[0058] Specifically, in step 2, the probability p of water vapor flux divergence at a specified time is calculated as follows:

[0059]

[0060] i y,m ∈[1, n m ];

[0061] Among them, VIMD y,m is the water vapor flux divergence in the mth month of the specified year y; m is the number of the month from January to December; p(VIMDy,m) is the probability of water vapor flux divergence in the mth month of the specified year y; n m = the number of samples of water vapor flux divergence in month m of different years in the multi-year period + 1; i y,m The definition is: the water vapor flux divergence data sets of the specified month m of year y and the water vapor flux divergence data sets of the month m of different years in the same period of multiple years are arranged in ascending order and assigned 1, 2, ..., up to n. m The serial number, i y,m is the sequence number of the water vapor flux divergence in the mth month of the yth year in the arrangement.

[0062] Compared with the prior art, in the present invention, when calculating the probability p of water vapor flux divergence at a specified time, the probability p of water vapor flux divergence at a specified time is calculated by calculating the probability p of water vapor flux divergence at a specified time. y,m The multi-year water vapor flux divergence in the relevant source historical data and the water vapor flux divergence at a specified time are sorted from small to large, covering all sample intervals of the source historical data. Compared with the partial sampling of historical data by the sampling method commonly used in the prior art, the present invention has a higher accuracy rate. At the same time, the present invention only mathematically sorts the sample intervals, does not require more complex calculations, and can be easily implemented with the assistance of a computer, thereby increasing the accuracy without excessively increasing the workload. At the same time, compared with another common method in the prior art that determines the specific distribution of probability through full mathematical analysis, the data processing workload is greatly simplified. By arranging and comparing the water vapor flux divergence of the mth month of the yth year and the water vapor flux divergence of the mth month of the multi-year period from small to large, and then calculating by a formula, the probability of the water vapor flux divergence of the specified mth month of the yth year can be obtained from the water vapor flux divergence.

[0063] Step 3: calculating the water vapor flux divergence index of the specified time according to the occurrence probability of the water vapor flux divergence of the specified time; normalizing the occurrence probability of the dispersedly distributed water vapor flux divergence into a normal distribution, and calculating the normal distribution index t of the water vapor flux divergence which is symmetrical to p=0.5 by using the empirical method of cumulative normal distribution approximation; further obtaining the water vapor flux divergence index SVIMDI by t through an empirical formula, wherein the SVIMDI is symmetrical to the point of SVIMDI=0, which is beneficial to further matching with the plant drought grade and realizing the quantitative rating of the plant drought grade.

[0064] Specifically, in step 3, the calculation method of the water vapor flux divergence index (SVIMDI) is as follows:

[0065] The occurrence probability p of the water vapor flux divergence of the specified time is divided into two cases of ≤0.5 and >0.5 to normalize the water vapor flux divergence into a normal distribution:

[0066] The water vapor flux divergence normal distribution index t is obtained according to the following formula:

[0067]

[0068] The water vapor flux divergence index SVIMDI obtained according to t further satisfies:

[0069]

[0070] Compared with the prior art, the water vapor flux divergence probability distribution of the present application is originally skewed distribution, in order to facilitate use and analysis, the probability of the water vapor flux divergence is converted into the water vapor flux divergence index SVIMDI by normalizing the distribution, and the occurrence probability p of the water vapor flux divergence is displayed symmetrically to the normal distribution, which is convenient for corresponding to the two states of the plant drought-wet and dry, and has universal significance. In the prior art, the similar drought index or occurrence probability is often asymmetric data, the data on one side is extremely dense, and the data on the other side is extremely dispersed, which cannot reflect the real situation of the occurrence probability of dry years and wet years. The water vapor flux divergence index SVIMDI of the present application can be displayed symmetrically to the normal distribution, which completely avoids the interference of the graphical display form on the effect display.

[0071] Step 4: Obtain the satellite remote sensing vegetation index of the target area, obtain the satellite remote sensing vegetation change index based on the satellite remote sensing vegetation index, perform linear regression statistical calculation based on the water vapor flux divergence index and satellite remote sensing vegetation change index at the specified time obtained in step 3 to obtain the significant areas of the water vapor flux divergence index and satellite remote sensing vegetation change index at the specified time in the target area; screen out the areas in the target area that have a significant correlation with the satellite remote sensing vegetation change index, and eliminate the interference of factors such as insect pests and extreme temperatures.

[0072] Specifically, in step 4, the method for obtaining the satellite remote sensing vegetation change index VA of the target area is:

[0073]

[0074] Among them, VA is the satellite remote sensing vegetation change index; y is the specified year, m is the number of January to December, VA y,m is the remote sensing vegetation change index for the designated month m of year y; VI y,m is the vegetation index for the mth month of the specified year y; VI y,m Products from NOAA, MODIS or domestic Fengyun satellites can be used, specifically Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Ratio Vegetation Index (RVI), Leaf Area Index (LAI), Photosynthetically Active Radiation (FPAR), Gross Primary Productivity (GPP) and Net Primary Productivity (NPP) indices that reflect vegetation growth and nutritional information; VI ave,m is the mean value of vegetation index in the mth month, VI ave,m It is calculated from the average vegetation index of the mth month over the years from products derived from the US NOAA, MODIS or domestic Fengyun satellites.

[0075] Specifically, in step 4, the linear regression statistical calculation method of the water vapor flux divergence index and the satellite remote sensing vegetation change index is as follows: using commonly used data analysis software, a linear regression model is constructed with the water vapor flux divergence index to be processed as the prediction factor and the satellite remote sensing vegetation change index as the prediction quantity, and the linear Pearson correlation coefficient is used to calculate the correlation; the water vapor flux divergence index data set and the corresponding satellite remote sensing vegetation change index data set are subjected to variance analysis to evaluate the F overall significance level of the regression model.

[0076] Furthermore, in step 4, the method for determining the significant region is: based on the overall significance test of variance analysis (F overall significance test), the region with a significance level less than 0.05 is selected as the significant region.

[0077] Specifically, if the correlation is positive and the significance level is less than 0.05, it means that the less water resources, the more damaged the vegetation; if the correlation is negative and the significance level is less than 0.05, it may be disturbed by floods. Therefore, in order to judge the situation of vegetation drought, in addition to judging the water vapor flux divergence index SVIMDI and remote sensing vegetation change index VA y,m In addition to significance, the correlation between the two needs to be determined.

[0078] To further improve assessment / forecast accuracy, the target area can be divided into several basic units from which the water vapor flux divergence index and / or satellite remote sensing vegetation change index can be obtained before the F-based significance test. The water vapor flux divergence index and satellite remote sensing vegetation change index of each basic unit are then used to perform the F-based significance test to determine whether the basic unit is a significant area. To the extent possible, the smaller the basic unit area, the more conducive it is to improving assessment / forecast accuracy. The size of the basic unit area from which the water vapor flux divergence index and / or satellite remote sensing vegetation change index can be obtained is limited by actual technical level and the ability to obtain meteorological and satellite remote sensing data.

[0079] Furthermore, a linear regression model is constructed using the water vapor flux divergence index of the basic unit identified as a significant area as a prediction factor and the satellite remote sensing vegetation change index of the basic unit as a prediction variable. Specifically, analysis software can directly perform linear regression analysis and construct a linear regression model. Compared to the drought indices commonly used in the prior art, the present invention, through a significance analysis of the water vapor flux divergence index and the satellite remote sensing vegetation change index in the target area, excludes areas in the target area where water vapor flux divergence changes but vegetation growth does not change significantly due to its strong drought tolerance. This avoids the one-sided description of drought from the perspective of atmospheric water shortage, while taking into account the different drought tolerance levels of different plants and the essential differences between plant drought and environmental drought, greatly improving the accuracy of plant drought assessment.

[0080] At the same time, compared with the existing method of using satellite remote sensing to monitor vegetation growth and then evaluate vegetation drought, satellite remote sensing monitoring of vegetation growth cannot eliminate the influence of factors such as insect pests and extreme temperatures on vegetation growth. y,m Significance analysis eliminates interference from factors such as insect pests and extreme temperatures. Furthermore, compared to existing methods that use satellite remote sensing to monitor vegetation growth and subsequently assess vegetation drought, the water vapor flux divergence index has a preemptive effect on vegetation drought. Changes in water vapor flux divergence primarily lead to changes in atmospheric moisture, requiring feedback from vegetation to regulate water before affecting vegetation growth. Therefore, the present invention uses water vapor flux divergence as a method for assessing vegetation drought, enabling prediction of vegetation growth within a region for a certain period of time.

[0081] Step 5: Classify the vegetation drought level of significant areas according to the water vapor flux divergence index at the specified time, and evaluate or predict the regional vegetation drought status.

[0082] Specifically, if SVIMDI>1.0, the target area or the basic unit of the target area is determined to be a wet area; if -1.0<SVIMDI≤1.0, the target area or the basic unit of the target area is determined to be a normal area; if SVIMDI≤-1.0, the target area or the basic unit of the target area is determined to be a drought area. The specific classification of drought intensity levels is shown in Table 1:

[0083] Table 1

[0084]

[0085] Compared with the existing technology, the present invention adopts SVIMDI and -1 and 1 as threshold endpoints, providing a clear and quantifiable evaluation standard for the complex and changeable vegetation drought affected by multiple factors, and proves in specific implementation examples that it has better prediction accuracy than the existing technology.

[0086] On the other hand, the present invention discloses a vegetation drought index assessment system based on water vapor flux divergence, which specifically includes the following modules:

[0087] Data acquisition and processing module: used to obtain historical data of water vapor flux divergence in the target area and convert it into source historical data;

[0088] Specifically, historical data on water vapor flux divergence in the target area can be obtained through meteorological data query; in order to improve the accuracy and reliability of the evaluation results, historical data from multiple years can be obtained. For example, the historical data can include data from the target area for more than 15 consecutive years, and the obtained historical data can be converted into source historical data that can be further processed, such as monthly average data: historical data can be monthly data, quarterly data, or short-term data with a timing period of days or even hours; monthly data and quarterly data can be directly used as source historical data; short-term data with a timing period of days or even hours can be converted into monthly average data or quarterly data using methods such as weighted averaging as needed.

[0089] Compared with the common drought index in existing technologies, the water vapor flux divergence represents the regional water vapor input and output, and uses a holistic approach to consider various factors affecting vegetation drought. This reduces the difficulty of combining variables such as precipitation and solar evaporation, which are mutually influential and selected in existing technologies, reduces the amount of data processing, and improves processing efficiency.

[0090] First calculation module:

[0091] The occurrence probability of the water vapor flux divergence at the specified time is calculated according to the source historical data;

[0092] The specified time is a time point at which the vegetation drought state of the target region is to be evaluated, and the specified time can be determined as a certain month of a certain year or a certain quarter of a certain year, for example, the yth year and the mth month in the historical data to be evaluated are selected as the specified time, and the water vapor flux divergence at the specified time and the water vapor flux divergence at the same period of the same month in multiple years are queried in the source historical data; the same period of multiple years refers to the time point in the historical year which is the same as the month of the specified time; specifically, the vegetation drought state of the target region in May 2019 is to be evaluated, and the data set of the source historical data of the water vapor flux divergence in May 1982-2019 is selected as the multiple-year same-period data of the water vapor flux divergence data in May 2019; specifically, the vegetation drought state of the target region in the first quarter of 2019 is to be evaluated, and the data set of the source historical data of the water vapor flux divergence in the first quarter of 1982-2019 is selected as the multiple-year same-period data of the water vapor flux divergence data in the first quarter of 2019.

[0093] Specifically, according to the source historical data of the water vapor flux divergence, the water vapor flux divergence at the specified time and the corresponding multiple-year same-period water vapor flux divergence are extracted, and the occurrence probability of the water vapor flux divergence at the specified time is calculated according to the two kinds of water vapor flux divergences.

[0094] Specifically, in step 2, the calculation method of the occurrence probability p of the water vapor flux divergence at the specified time is as follows:

[0095]

[0096] i y,m ∈[1,n m ];

[0097] Wherein, VIMD y,m is the water vapor flux divergence of the specified yth year and mth month; m is the number of 1-12 months; p(VIMDy,m) is the probability of the water vapor flux divergence of the specified yth year and mth month; n m = the sample number of the water vapor flux divergence of the mth month in different years in the multiple-year same period + 1; i y,m is defined as follows: the data set of the water vapor flux divergence of the specified yth year and mth month and the water vapor flux divergence of the mth month in different years in the multiple-year same period is arranged in ascending order according to the numerical value, and is sequentially assigned with 1, 2, …, and n m , i y,m is the sequence number of the water vapor flux divergence of the yth year and mth month in the arrangement.

[0098] Compared with the prior art, in the present application, when calculating the occurrence probability p of the water vapor flux divergence at the specified time, the i y,mThe multi-year water vapor flux divergence in the relevant source historical data and the water vapor flux divergence at a specified time are sorted from small to large, covering all sample intervals of the source historical data. Compared with the partial sampling of historical data by the sampling method commonly used in the prior art, the present invention has a higher accuracy rate. At the same time, the present invention only mathematically sorts the sample intervals, does not require more complex calculations, and can be easily implemented with the assistance of a computer, thereby increasing the accuracy without excessively increasing the workload. At the same time, compared with another common method in the prior art that determines the specific distribution of probability through full mathematical analysis, the data processing workload is greatly simplified. By arranging and comparing the water vapor flux divergence of the mth month of the yth year and the water vapor flux divergence of the mth month of the multi-year period from small to large, and then calculating by a formula, the probability of the water vapor flux divergence of the specified mth month of the yth year can be obtained from the water vapor flux divergence.

[0099] Second calculation module:

[0100] Calculate the water vapor flux divergence index at a specified time according to the occurrence probability of water vapor flux divergence at a specified time;

[0101] The probability of occurrence of dispersed water vapor flux divergence is normalized and calculated using an empirical method that approximates the cumulative normal distribution: the probability of occurrence of water vapor flux divergence at a specified time, p, is converted into a normal distribution standardized index t of water vapor flux divergence symmetrical with the variable p = 0.5; t is further converted into the water vapor flux divergence index SVIMDI through an empirical formula, where SVIMDI is symmetrically distributed around the point SVIMDI = 0, which is conducive to further matching with the plant drought disaster level and realizing quantitative rating of the plant drought disaster level.

[0102] Specifically, the calculation method of the water vapor flux divergence index (SVIMDI) is:

[0103] The probability p of water vapor flux divergence at a specified time is divided into two cases: ≤0.5 and >0.5, and the water vapor flux divergence is normalized to normal distribution:

[0104] The normalized index t of the water vapor flux divergence is obtained according to the following formula:

[0105]

[0106] According to t, the water vapor flux divergence index SVIMDI satisfies:

[0107]

[0108] Compared with the prior art, the water vapor flux divergence probability distribution of the present application is originally skewed distribution, in order to facilitate use and analysis, the probability of water vapor flux divergence is converted into water vapor flux divergence index SVIMDI through standard normal distribution, the occurrence probability p of water vapor flux divergence is symmetrically displayed in normal distribution, which is convenient for corresponding to two states of plant drought and wetness, and has universal significance. In the prior art, the calculation of similar drought index or occurrence probability is often asymmetric data, the data on one side is extremely dense, and the data on the other side is extremely discrete, and the value of the index cannot intuitively reflect the real situation of the occurrence probability of dry years and wet years. The water vapor flux divergence index SVIMDI of the present application is symmetrically displayed in normal distribution, and the statistical data rule can be intuitively understood from the value of the index.

[0109] The third calculation module comprises:

[0110] The satellite remote sensing vegetation index of the target region is obtained, the satellite remote sensing vegetation change index is obtained according to the satellite remote sensing vegetation index, and the significance region of the water vapor flux divergence index and the satellite remote sensing vegetation change index of the target region at the specified time is obtained by regression statistics according to the water vapor flux divergence index and the satellite remote sensing vegetation change index at the specified time obtained in the above steps; the region having significant correlation with the satellite remote sensing vegetation change index in the target region is screened out by fully considering the spatio-temporal heterogeneity of the response of vegetation to drought.

[0111] Specifically, the method for obtaining the satellite remote sensing vegetation change index VA of the target region comprises:

[0112]

[0113] Wherein, VA is the satellite remote sensing vegetation change index; y is a specified year, m is the number of 1-12 months, VA y,m is the remote sensing vegetation change index in the specified yth year and mth month; VI y,m is the vegetation index in the specified yth year and mth month; VI y,m may be derived from the products of NOAA, MODIS or domestic Fengyun satellites, and specifically can be normalized vegetation index (NDVI), enhanced vegetation index (EVI), ratio vegetation index (RVI), leaf area index (LAI), photosynthetically active radiation component (FPAR), gross primary productivity (GPP) and net primary productivity (NPP) index reflecting vegetation growth and nutrition information; VI ave,m is the mean value of the vegetation index of the mth month in the calendar year, VI ave,m is calculated from the mean value of the vegetation index of the mth month in the calendar year derived from the products of NOAA, MODIS or domestic Fengyun satellites.

[0114] Furthermore, the method for determining the significant region is: constructing a regression model, and screening the region with a significance level less than 0.05 as the significant region according to the overall significance test of variance analysis (F overall significance test).

[0115] Specifically, if the correlation is positive and the significance level is less than 0.05, it means that the less water resources, the more damaged the vegetation; if the correlation is negative and the significance level is less than 0.05, it may be affected by floods. Therefore, in order to judge the situation of vegetation drought, in addition to using the water vapor flux divergence index SVIMDI and the remote sensing vegetation change index VA y,m In addition to significance, the correlation between the two needs to be determined.

[0116] Specifically, the linear regression statistical method of the water vapor flux divergence index and the satellite remote sensing vegetation change index is as follows: using commonly used data analysis software, a linear regression model is constructed with the water vapor flux divergence index as the prediction factor and the satellite remote sensing vegetation change index as the prediction quantity, the linear Pearson correlation coefficient is used to calculate the correlation, and the variance analysis is performed on the water vapor flux divergence index dataset to be processed and the corresponding satellite remote sensing vegetation change index dataset to evaluate the F overall significance level of the regression model.

[0117] To further improve assessment / forecast accuracy, before performing the F-based significance test, the target area can be divided into several basic units from which the water vapor flux divergence index and / or satellite remote sensing vegetation change index can be obtained. The F-based significance test is then performed using the water vapor flux divergence index and satellite remote sensing vegetation change index of each basic unit to determine whether the basic unit is a significant area. To the extent possible, smaller basic unit areas are more conducive to improved assessment / forecast accuracy. However, the size of the basic unit for obtaining the water vapor flux divergence index and / or satellite remote sensing vegetation change index is limited by practical technical capabilities and the ability to obtain meteorological and satellite remote sensing data.

[0118] Furthermore, a linear regression model is constructed using the water vapor flux divergence index of the basic units identified as significant areas as a predictor and the satellite remote sensing vegetation change index of the basic units as a predictor. Compared to the drought indices commonly used in the prior art, this method, through a significance analysis of the water vapor flux divergence index and the satellite remote sensing vegetation change index within the target area, excludes areas in the target area where water vapor flux divergence changes but vegetation growth does not change significantly due to its strong drought tolerance. This avoids the one-sided description of drought from the perspective of atmospheric water shortage, while also taking into account the different tolerance levels of plants to drought, greatly improving the accuracy of plant drought assessment.

[0119] At the same time, compared with the existing method of using satellite remote sensing to monitor vegetation growth and then evaluate vegetation drought, satellite remote sensing monitoring of vegetation growth cannot eliminate the influence of factors such as insect pests and extreme temperatures on vegetation growth. y,m Significance analysis excluded areas unaffected by changes in water vapor flux divergence. Compared to existing methods that use satellite remote sensing to monitor vegetation growth and subsequently assess vegetation drought, the water vapor flux divergence index has a preemptive effect on vegetation drought. Changes in water vapor flux divergence primarily lead to changes in atmospheric moisture, requiring feedback from vegetation to regulate moisture before affecting vegetation growth. Therefore, using water vapor flux divergence as a vegetation drought assessment method in this invention can predict vegetation growth within a region for a certain period of time in the future.

[0120] Drought level determination module:

[0121] Based on the water vapor flux divergence index at a specified time, the vegetation drought level of significant areas is divided into different levels to evaluate or predict the regional vegetation drought conditions.

[0122] Specifically, if SVIMDI>1.0, the target area or the basic unit of the target area is determined to be a wet area; if -1.0<SVIMDI≤1.0, the target area or the basic unit of the target area is determined to be a normal area; if SVIMDI≤-1.0, the target area or the basic unit of the target area is determined to be an arid area.

[0123] In order to further describe the vegetation drought index assessment method and system based on water vapor flux divergence in detail, examples and comparative examples are further set up:

[0124] Example

[0125] Based on the previously disclosed implementation methods, this embodiment discloses a method and system for evaluating vegetation drought index based on water vapor flux divergence:

[0126] First, the vertically integrated water vapor divergence dataset (VIMD) from the global 0.25-degree atmospheric reanalysis data provided by the European Centre for Medium-Range Weather Forecasts from 1982 to 2019 was preprocessed into a monthly-scale dataset with 0.25-degree as the basic unit. A water vapor flux divergence dataset was extracted based on the boundary of the Tarim River Basin. As an example, the basic unit (77.125°E-77.375°E, 37.625°N-37.875°N) in the 1983 VIMD monthly-scale dataset is selected as shown in Table 2.

[0127] Secondly, the occurrence probability p of water vapor flux divergence is calculated according to the following formula:

[0128]

[0129] Among them, the basic unit (77.125°E-77.375°E, 37.625°N-37.875°N) is selected as an example, p(VIMD y,m ) as the probability of water vapor flux divergence in the region from May to September 1983. The specific results are shown in Table 2.

[0130] Furthermore, the water vapor flux divergence was normalized based on the probability of ≤0.5 and >0.5. The target basin was divided into basic units, and the SVIMDI of the basin-wide grid was calculated. As an example, the basic unit (77.125°E-77.375°E, 37.625°N-37.875°N) was selected to calculate the SVIMDI of this area from May to September 1983. The specific results are shown in Table 2.

[0131] Furthermore, the NOAA satellite Blended-VHP 1982-2019 4km vegetation NDVI was resampled to 0.25 degrees, and the VA of NDVI was calculated according to the following formula: y,m Vegetation Change Index:

[0132]

[0133] VA y,m is the remote sensing vegetation change index for month m of year y; VI y,m is the vegetation index for month m of year y; VI ave,m is the mean value of the vegetation index in the mth month over the years.

[0134] Furthermore, the monthly SVIMDI is used as the prediction factor for each grid, and VA y,m A linear regression model was constructed to predict the amount of vegetation change in the Tarim River Basin during the growing season (May-September) and the average correlation between SVIMDI and the vegetation change in the optical remote sensing area was 0.238. y,m It is positively correlated; according to the F overall significance test, it is judged whether the prediction model can pass the 0.05 significance level test. y,m The region tested at the 0.05 significance level is as follows Figure 2 As shown: In 1983, the vegetation changes in the growing season (May-September) of the region from May to September based on optical remote sensing and the basic units that passed the 0.05 significance level test of SVIMDI are distributed in dark solid color.

[0135] As an example, the basic unit (77.125°E-77.375°E, 37.625°N-37.875°N) is selected to calculate the SVIMDI and VA of this area from May to September from 1982 to 2019. y,m The significance level of:

[0136] Regression equation: VA = 0.048 × (SVIMDI) - 0.027;

[0137] F value: 24.957;

[0138] Significance: 0.00;

[0139] Pearson correlation coefficient: 0.3439.

[0140] Table 2

[0141]

[0142] The grid area where vegetation was significantly affected by drought during the growing season (May-September) in the Tarim River Basin optical remote sensing region was 99,043 square kilometers. The Pearson correlation coefficient was greater than 0, indicating that SVIMDI and VA y,m The F overall significance test was less than 0.05 to determine whether SVIMDI and VA y,m According to the disaster records of Xinjiang Kizilsu Kirgiz Autonomous Prefecture in the Tarim River Basin from 1982 to 2000 in the China Meteorological Disaster Dictionary, the vegetation drought disaster levels are as follows based on the annual average SVIMDI changes in the autonomous prefecture: Figure 3 As shown in the figure: the vegetation drought disaster level is judged by the SVIMDI value; if SVIMDI>1.0, the target area or the unit of the target area is judged to be a wet area at the specified time; if -1.0<SVIMDI≤1.0, the target area or the unit of the target area is judged to be a normal area at the specified time; if SVIMDI≤-1.0, the target area or the unit of the target area is judged to be a drought area at the specified time.

[0143] As an example, as shown in Table 3, 1997, 1991, and 1983 were drought years. As shown in Table 4, the annual average SVIMDI of the autonomous prefecture was -7.28, -6.06, and -2.50, respectively. The drought level was divided into -1.0 as the threshold, which is consistent with the local drought disaster records. Therefore, SVIMDI and thresholds can be applied to oasis agricultural areas. y,m In the significantly correlated areas, the vegetation drought level was delineated using the SVIMDI thresholds of -1.0 and 1.0.

[0144] Table 3

[0145]

[0146] Table 4

[0147]

[0148] Comparative Example

[0149] The vertically integrated water vapor divergence dataset (VIMD) in the global 0.25-degree atmospheric reanalysis data from 1982 to 2019 provided by the European Centre for Medium-Range Weather Forecasts was preprocessed into a monthly scale dataset with 0.25-degree as the basic unit. The standardized precipitation index (SPI) monthly drought index was calculated using the existing technology (Mckee TB, Doesken NJ, Kleist J. The Relationship of Drought Frequency and Duration to Time Scales. Eighth Conference on Applied Climatology, 1993: 179-184). Under the same conditions as in the embodiment, the monthly SPI was used as the prediction factor for each grid instead of the SVIMDI in the embodiment, and the VA was used. y,m A linear regression model was constructed for the prediction and an F-value test was performed to determine the vegetation change VA in the growing season (May-September) of the Tarim River Basin using optical remote sensing. y,m The average correlation with SPI is 0.202. y,m It is positively correlated; according to the F overall significance test, it is judged whether the prediction model can pass the 0.05 significance level test. y,m The region tested at the 0.05 significance level is as follows Figure 4 As shown in the figure: In 1983, the vegetation changes in the region during the growing season (May-September) from optical remote sensing in this region and the basic units that passed the SPI test at the 0.05 significance level are distributed in dark solid colors; the F overall significance test shows that the grid area significantly affected by drought (significance level less than 0.05) is 15,756.9 square kilometers. The vegetation drought disaster level is determined based on the annual average SPI change in the autonomous prefecture: SPI>1.0, the target area or the basic unit of the target area is judged to be in the wet zone at the specified time; -1.0<SPI≤1.0, the target area or the basic unit of the target area is judged to be in the normal zone at the specified time; SPI≤-1.0, the target area or the basic unit of the target area is judged to be in the drought zone at the specified time. According to the disaster records of Xinjiang Volume, China Meteorological Disasters Encyclopedia, Kizilsu Kirgiz Autonomous Prefecture, Xinjiang from 1982 to 2000, 1997, 1991, and 1983 were drought years. Figure 5As shown, the annual average SPI for the autonomous prefecture in 1997, 1991, and 1983 were 2.197, 10.165, and -9.151, respectively. Using a threshold of -1.0 for drought classification, these values ​​differ significantly from local drought records. Therefore, the SPI index is not suitable for predicting vegetation drought in this region.

[0150] It can be seen that the SVIMDI vegetation drought prediction method in the embodiment has wider applicability and accuracy in the Tarim River Basin, especially in Kizilsu Kirgiz Autonomous Prefecture, compared with the method in the comparative example under the same conditions.

[0151] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed by the present invention should be covered by the scope of protection of the present invention.

Claims

1. The vegetation drought index assessment method based on water vapor flux divergence is characterized by: The following steps are involved: Acquire historical data of water vapor flux divergence of a target area, and obtain source historical data based on the historical data; the historical data includes water vapor flux divergence data of the target area obtained by querying meteorological data; the source historical data includes water vapor flux divergence data of a required time unit; The probability of occurrence of water vapor flux divergence at a specified time is calculated based on historical source data; Calculate the water vapor flux divergence index at a specified time according to the occurrence probability of water vapor flux divergence at a specified time; Obtain the satellite remote sensing vegetation index of the target area, obtain the satellite remote sensing vegetation change index based on the satellite remote sensing vegetation index, and perform linear regression analysis based on the obtained water vapor flux divergence index and satellite remote sensing vegetation change index at the specified time to obtain the significant area of ​​the water vapor flux divergence index and satellite remote sensing vegetation change index at the specified time in the target area; The vegetation drought level of significant areas is divided according to the water vapor flux divergence index at a specified time.

2. The vegetation drought index evaluation method according to claim 1, characterized in that: The calculation method of the occurrence probability p of the water vapor flux divergence at the specified time is: i y,m ∈[1,n m ]; Among them, VIMD y,m is the water vapor flux divergence in the mth month of the specified year y; m is the number of the month from January to December; p(VIMDy,m) is the probability of water vapor flux divergence in the mth month of the specified year y; n m = the number of samples of water vapor flux divergence in month m of different years in the multi-year period + 1; i y,m The definition is: the water vapor flux divergence data sets of the specified month m of year y and the water vapor flux divergence data sets of the month m of different years in the same period of multiple years are arranged in ascending order and assigned 1, 2, ..., up to n. m The serial number, i y,m is the sequence number of the water vapor flux divergence in the mth month of the yth year in the arrangement.

3. The vegetation drought index evaluation method according to claim 1, characterized in that: The calculation method of the water vapor flux divergence index at the specified time is: The occurrence probability p of water vapor flux divergence VIMD is divided into two cases ≤ 0.5 and > 0.5, and the water vapor flux divergence is normalized to obtain the normal distribution normalization index t of water vapor flux divergence, which satisfies: The water vapor flux divergence index SVIMDI satisfies:

4. The vegetation drought index evaluation method according to claim 1, characterized in that: The method for obtaining the satellite remote sensing vegetation change index VA of the target area is: y is the year, m is the number of the month from January to December, VA y,m is the remote sensing vegetation change index for the mth month of the specified yth year; VI y,m is the vegetation index for the mth month of the specified year y; VI ave,m is the mean value of the vegetation index in the mth month over the years.

5. The vegetation drought index evaluation method according to claim 4, characterized in that: The method for determining the significant areas of the water vapor flux divergence index and the satellite remote sensing vegetation change index is as follows: performing linear regression analysis on the water vapor flux divergence index and the satellite remote sensing vegetation change index, and screening the areas with a significance level less than 0.05 according to the F overall significance test to determine them as significant areas.

6. The vegetation drought index evaluation method according to claim 5, characterized in that: Before the F overall significance test, the target area is divided into a number of basic units from which the water vapor flux divergence index and / or the satellite remote sensing vegetation change index can be obtained.

7. The vegetation drought index evaluation method according to claim 6, characterized in that: A linear regression model was constructed with the water vapor flux divergence index of the basic unit as the prediction factor and the satellite remote sensing vegetation change index of the basic unit as the prediction quantity, and the areas with a significance level less than 0.05 were screened as significant areas according to the F overall significance test.

8. The vegetation drought index evaluation method according to claim 7, characterized in that: The classification standard of the vegetation drought disaster level is: SVIMDI>1.0, then the target area or the basic unit of the target area is determined to be a wet area; -1.0<SVIMDI≤1.0, then the target area or the basic unit of the target area is determined to be a normal area; SVIMDI≤-1.0, then the target area or the basic unit of the target area is determined to be a drought area.

9. The vegetation drought index assessment method according to any one of claims 1 to 8, characterized in that: The historical data includes at least 15 years of water vapor flux divergence data.

10. The vegetation drought index assessment system based on water vapor flux divergence is characterized by: Used to implement the evaluation method according to any one of claims 1 to 9, comprising the following modules: Data acquisition and processing module: Used to obtain historical data of water vapor flux divergence in the target area, and obtain source historical data based on the historical data processing; the historical data includes water vapor flux divergence data of the target area obtained by querying meteorological data; the source historical data includes water vapor flux divergence data of the required time unit; First calculation module: The probability of occurrence of water vapor flux divergence at a specified time is calculated based on historical source data; Second calculation module: Calculate the water vapor flux divergence index at a specified time according to the occurrence probability of water vapor flux divergence at a specified time; The third calculation module: Obtain the satellite remote sensing vegetation index of the target area, obtain the satellite remote sensing vegetation change index based on the satellite remote sensing vegetation index, and perform linear regression analysis based on the obtained water vapor flux divergence index and satellite remote sensing vegetation change index at the specified time to obtain the significant area of ​​the water vapor flux divergence index and satellite remote sensing vegetation change index at the specified time in the target area; Drought level determination module: The vegetation drought level of significant areas is divided according to the water vapor flux divergence index at a specified time.

Images