A meteorological-vegetation drought transmission characteristic identification method and system based on a three-dimensional clustering algorithm

Abstract

The application discloses a meteorological-vegetation drought transmission characteristic identification method and system based on a three-dimensional clustering algorithm, which comprises the following steps: S1, identifying meteorological drought events and vegetation drought events in a research period based on a three-dimensional clustering algorithm; S2, determining characteristic variables of the meteorological drought events and the vegetation drought events; S3, proposing a drought transmission identification rule based on three-dimensional space-time clustering, and matching meteorological-vegetation drought event pairs; S4, determining types of the meteorological-vegetation drought event pairs; and S5, extracting characteristic variables of a meteorological drought to vegetation drought transmission process, and clarifying the meteorological-vegetation drought transmission process. The method and system can quantitatively identify characteristic changes in a meteorological-vegetation drought transmission process, reveal transmission characteristics of meteorological drought to vegetation drought, clarify the internal relationship, response characteristics and transformation mechanism between meteorological drought and vegetation drought, and provide scientific and technological support for accurately predicting and preventing vegetation drought.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of drought transmission recognition, and in particular to a meteorological-vegetation drought transmission feature recognition method and system based on a three-dimensional clustering algorithm. BACKGROUND

[0002] At present, the recognition of drought events at home and abroad mostly adopts a one-dimensional time recognition method (run theory) to extract drought duration, intensity, peak value and other information from a time series of drought indicators by setting a threshold. The run theory method recognizes drought events from the time dimension, but it reduces drought events to a low-order subspace and loses a large amount of spatiotemporal information, so it cannot comprehensively reflect the change characteristics of drought events in the time dimension and the spatial dimension. Andreadis proposed a clustering method using a severity-duration-area curve to extract the spatial characteristics of drought, and analyzed the evolution trend and law of drought by analyzing the spatial distribution of drought in different periods. Lloyd-Hughes further improved the drought event recognition method to the time-latitude-longitude dimension on the basis of previous research. Xu Kai et al. further improved the three-dimensional drought event recognition method based on three-dimensional spatiotemporal clustering.

[0003] Different types of droughts have a correlation relationship, and the matching of drought events is often used to analyze the response relationship between different types of drought events. A drought event is a three-dimensional drought body integrating time and space, and the transmission of a drought event is a three-dimensional dynamic process in time and space. At present, there are few studies on the transmission process of drought events from a three-dimensional perspective. Liu Yi et al. developed a three-dimensional drought event matching method based on time-latitude-longitude, which comprehensively considered the three-dimensional overlap of drought events in terms of time duration and spatial coverage. Feng Kai et al. used the three-dimensional drought event matching method to match the meteorological-vegetation drought events in the Heihe River Basin in China from 1961 to 2014, and the results showed that this method could improve the accuracy of matching meteorological-vegetation drought events.

[0004] However, due to the complexity of the formation and evolution of drought, previous three-dimensional drought event matching methods mostly qualitatively evaluate the transmission process of drought, which cannot accurately and truly reflect the transmission process of drought, and has great limitations in identifying the transmission characteristics of drought. Therefore, matching the characteristics of drought events in the time and spatial dimensions and quantitatively identifying the transmission characteristics of drought are key problems to be solved for clarifying the transmission mechanism between droughts. SUMMARY

[0005] The present application relates to the technical field of drought transmission recognition, and in particular to a meteorological-vegetation drought transmission feature recognition method and system based on a three-dimensional clustering algorithm.

[0006] To achieve the above-mentioned purpose, the technical solution adopted by the present application is as follows:

[0007] A meteorological-vegetation drought transmission characteristic identification method based on a three-dimensional clustering algorithm, comprising the following steps,

[0008] S1, drought event identification:

[0009] Based on the three-dimensional clustering algorithm, the drought patch is determined according to the drought index size of meteorological drought and vegetation drought, and all meteorological drought events and vegetation drought events in the research period are identified according to the drought patch area and the size of the overlapping area of the drought patch in adjacent months;

[0010] S2, drought event analysis:

[0011] The dynamic changes of the identified meteorological drought events and vegetation drought events are analyzed from multiple drought event characteristic variables, and the characteristic variables of the corresponding drought events are determined;

[0012] S3, drought event pairing:

[0013] Based on the drought transmission identification rule of three-dimensional space-time clustering, the meteorological drought events and vegetation drought events are matched according to the overlapping time and overlapping area size in the time and space dimensions of the meteorological drought events and vegetation drought events, and a meteorological-vegetation drought event pair is obtained.

[0014] S4, event pair transmission type determination:

[0015] According to the number of meteorological drought events and the number of vegetation drought events, the drought transmission type of the meteorological-vegetation drought event pair is divided into one-to-one, one-to-many, many-to-one and many-to-many;

[0016] S5, drought transmission process analysis:

[0017] Based on the characteristic variables of the meteorological drought to vegetation drought transmission process, the transmission direction, end order, transmission type and drought occurrence range shrinkage trend of meteorological-vegetation drought are analyzed to clarify the meteorological-vegetation drought transmission process.

[0018] Preferably, step S1 specifically comprises the following contents,

[0019] S11, determining the drought index and index threshold of meteorological drought and vegetation drought;

[0020] S12, screening out the grid with drought index less than the corresponding index threshold at the monthly scale to form a drought patch;

[0021] S13, judging whether the drought patch area is less than the patch area threshold, if yes, considering that the drought patch is not established, and it is discarded; otherwise, it is retained and enters S14;

[0022] S14, calculate the overlapping area between adjacent drought patches of months, judge whether the overlapping area is less than the patch overlapping area threshold value, if yes, mark the two drought patches as two independent drought events, otherwise, mark the two drought patches as the same drought event;

[0023] S15, identify all drought events in the research period, delete the drought events with a drought event duration less than the duration threshold value, and number the remaining drought events to obtain meteorological drought events and vegetation drought events.

[0024] Preferably, the drought index of the meteorological drought is the standardized precipitation evapotranspiration index, and the index threshold value is-1; the drought index of the vegetation drought is the vegetation health index, and the index threshold value is 0.4.

[0025] Preferably, step S2 specifically comprises: analyzing the dynamic changes of meteorological drought events and vegetation drought events from five drought event characteristic variables, i.e., drought duration, influence area, intensity, strength and centroid, so as to determine the characteristic variables of the corresponding drought events.

[0026] Preferably, step S3 specifically comprises the following contents,

[0027] S31, sort the meteorological drought events and the vegetation drought events in time sequence;

[0028] S32, judge whether there is an intersection between the meteorological drought events and the vegetation drought events in time and space dimensions, when the overlapping time is greater than the overlapping time threshold value and the overlapping area is greater than the dimension overlapping area threshold value, it is considered that the meteorological-vegetation event pair is established, otherwise, it is not established;

[0029] S33, number the paired meteorological-vegetation event pairs in time sequence.

[0030] Preferably, the overlapping time threshold value is 0.

[0031] Preferably, step S4 specifically comprises that the drought transmission types of the meteorological-vegetation drought event pairs are divided into four types according to the number of meteorological drought events and the number of vegetation drought events, which are,

[0032] One-to-one: one meteorological drought event causes one vegetation drought event;

[0033] One-to-many: one meteorological drought event causes multiple vegetation drought events;

[0034] Many-to-one: multiple meteorological drought events cause one vegetation drought event;

[0035] Many-to-many: multiple meteorological drought events cause multiple vegetation drought events.

[0036] Preferably, step S5 specifically comprises the following contents,

[0037] When the difference between the starting time of the vegetation drought event and the starting time of the meteorological drought event is positive, it indicates that the meteorological-vegetation drought is positive transmission; otherwise, it indicates that the meteorological-vegetation drought event is negative transmission.

[0038] When the difference between the ending time of the vegetation drought event and the ending time of the meteorological drought event is positive, it indicates that the vegetation drought ends later than the meteorological drought; otherwise, it indicates that the vegetation drought ends earlier than the meteorological drought.

[0039] When the number of the meteorological drought events is equal to 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is one-to-one or one-to-many; when the number of the meteorological drought events is greater than 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is many-to-one or many-to-many.

[0040] When the number of the vegetation drought events is equal to 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is one-to-one or many-to-one; when the number of the vegetation drought events is greater than 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is one-to-many or many-to-many.

[0041] When the difference between the projection area of the meteorological drought and the projection area of the vegetation drought on the geographical coordinate plane is positive, it indicates that the occurrence range of the meteorological-vegetation drought shows a shrinking trend; otherwise, it indicates that the occurrence range of the meteorological-vegetation drought shows an expanding trend.

[0042] The present application also aims to provide a meteorological-vegetation drought transmission feature recognition system based on a three-dimensional clustering algorithm, which is used to implement any of the above-mentioned methods, and comprises,

[0043] The drought event recognition module: based on the three-dimensional clustering algorithm, the drought patches are determined according to the drought index size of the meteorological drought and the vegetation drought, and all meteorological drought events and vegetation drought events in the research period are recognized according to the drought patch area and the size of the overlapping area of the drought patch in adjacent months.

[0044] The drought event analysis module: the dynamic changes of the identified meteorological drought events and vegetation drought events are analyzed from multiple drought event feature variables, respectively, to determine the feature variables of the corresponding drought events.

[0045] The drought event pairing: based on the drought transmission recognition rules of the three-dimensional spatiotemporal clustering, the meteorological drought events and the vegetation drought events are matched according to the overlapping time and the size of the overlapping area of the meteorological drought events and the vegetation drought events in the time and space dimensions, to obtain the meteorological-vegetation drought event pair.

[0046] Event pair transmission type determination module: according to the number of meteorological drought events and the number of vegetation drought events, the drought transmission type of the meteorological-vegetation drought event pair is divided into one-to-one, one-to-many, many-to-one and many-to-many four types;

[0047] Drought transmission process analysis module: based on the characteristic variables of the meteorological drought to vegetation drought transmission process, the transmission direction, ending order, transmission type and drought occurrence range expansion trend of meteorological-vegetation drought are analyzed to clarify the meteorological-vegetation drought transmission process.

[0048] The beneficial effects of the present application are: 1. The method and system can match the characteristics of drought events in time and space dimensions, quantitatively identify the characteristic changes in the meteorological-vegetation drought transmission process, reveal the transmission characteristics of meteorological drought to vegetation drought, and clarify the transmission mechanism of meteorological drought to vegetation drought. 2. The method and system can quantitatively identify the characteristic changes in the meteorological-vegetation drought transmission process, reveal the transmission characteristics of meteorological drought to vegetation drought, clarify the internal relationship, response characteristics and transformation mechanism between meteorological drought and vegetation drought, and provide scientific and technological support for accurate prediction and prevention of vegetation drought. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 is a flowchart of the identification method in the embodiment of the present application;

[0050] Figure 2 is a schematic diagram of the dynamic change of a typical vegetation drought event in the embodiment of the present application;

[0051] Figure 3 is a transmission process diagram of a typical meteorological-vegetation drought event pair of the "one-to-one" type in the embodiment of the present application;

[0052] Figure 4 is a transmission process diagram of a typical meteorological-vegetation drought event pair of the "one-to-many" type in the embodiment of the present application;

[0053] Figure 5 is a transmission process diagram of a typical meteorological-vegetation drought event pair of the "many-to-one" type in the embodiment of the present application;

[0054] Figure 6 is a transmission process diagram of a typical meteorological-vegetation drought event pair of the "many-to-many" type in the embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0056] Embodiment one

[0057] In this embodiment, a meteorological-vegetation drought transmission characteristic identification method based on a three-dimensional clustering algorithm is provided, which includes five steps, specifically as follows.

[0058] I. Drought event identification:

[0059] Based on the three-dimensional clustering algorithm, the drought patches are determined according to the drought index size of meteorological drought and vegetation drought, and all meteorological drought events and vegetation drought events in the study period are identified according to the drought patch area and the size of the overlapping area of the drought patch in adjacent months.

[0060] Specifically, this step includes the following contents,

[0061] 1. Determine the drought index and index threshold of meteorological drought and vegetation drought; the drought index of meteorological drought is SPEI (i.e. standardized precipitation evapotranspiration index), and the index threshold is -1; the drought index of vegetation drought is VHI (i.e. vegetation health index), and the index threshold is 0.4.

[0062] 2. Screen out the grid with drought index less than the corresponding index threshold at the monthly scale to form a drought patch.

[0063] 3. Determine whether the drought patch area is less than the patch area threshold, if yes, it is considered that the drought patch is not established, and it is discarded; otherwise, it is retained and enters S14. The patch area threshold can be set according to actual conditions to better meet the needs.

[0064] 4. Calculate the overlapping area between adjacent month drought patches, and determine whether the overlapping area is less than the patch overlapping area threshold, if yes, mark the two drought patches as two independent drought events, otherwise, mark the two drought patches as the same drought event. The patch overlapping area threshold can be set according to actual conditions to better meet the needs.

[0065] 5. Identify all drought events in the study period, delete the drought events with a duration less than the duration threshold, and number the remaining drought events to obtain meteorological drought events and vegetation drought events.

[0066] II. Drought event analysis:

[0067] The dynamic changes of the identified meteorological drought events and vegetation drought events are analyzed from multiple drought event characteristic variables to determine the characteristic variables of the corresponding drought events.

[0068] Specifically, this step is to analyze the dynamic changes of meteorological drought events and vegetation drought events from five drought event characteristic variables, i.e. drought duration, influence area, intensity, strength, and centroid, to determine the characteristic variables of the corresponding drought events.

[0069] 1, drought duration is the length of time from the beginning to the end of each drought event, that is, the height of the three-dimensional drought body in the time dimension. Unit: month.

[0070] 2, the impact area is the area where each drought event occurs, that is, the maximum projection of the three-dimensional drought body on the spatial plane. Unit: km 2 .

[0071] 3, intensity is the sum of the water shortage degree of all drought grids, that is, the water shortage volume of the three-dimensional drought body. Unit: month·km 2 .

[0072] 4, intensity is the ratio between drought intensity and duration and impact area.

[0073] 5, the centroid is the position of each drought event in the latitude-longitude-time three-dimensional space, that is, the center of gravity of the three-dimensional drought body.

[0074] III. Drought event pairing:

[0075] Based on the drought transmission identification rule of three-dimensional space-time clustering, according to the overlapping time and the size of the dimension overlap area of the meteorological drought time and the vegetation drought event in time and space dimensions, the meteorological drought event and the vegetation drought event are matched, and the meteorological-vegetation drought event pair is obtained.

[0076] Specifically, the step includes the following contents,

[0077] 1, the meteorological drought event and the vegetation drought event are sorted in time sequence.

[0078] 2, judge whether there is intersection between the meteorological drought and the vegetation drought event in time and space dimensions, when the overlapping time is greater than the overlapping time threshold and the overlapping area is greater than the dimension overlap area threshold, it is considered that the meteorological-vegetation event pair is established, otherwise, it is not established. The overlapping time threshold and the dimension overlap area threshold can be set according to the actual situation, so as to meet the needs, in this embodiment, the overlapping time threshold is set to 0.

[0079] 3, the meteorological-vegetation event pair that is successfully paired is numbered in time sequence.

[0080] IV. Event pair transmission type determination:

[0081] According to the number of meteorological drought events and the number of vegetation drought events, the drought transmission type of the meteorological-vegetation drought event pair is divided into "one-to-one", "one-to-many", "many-to-one" and "many-to-many".

[0082] Specifically, the drought transmission types of the meteorological-vegetation drought event pairs are classified into four categories according to the number of meteorological drought events and the number of vegetation drought events, respectively,

[0083] 1. One-to-one: one meteorological drought event causes one vegetation drought event;

[0084] 2. One-to-many: one meteorological drought event causes multiple vegetation drought events;

[0085] 3. Many-to-one: multiple meteorological drought events cause one vegetation drought event;

[0086] 4. Many-to-many: multiple meteorological drought events cause multiple vegetation drought events.

[0087] Five, drought transmission process analysis:

[0088] Based on the characteristic variables of the meteorological drought to vegetation drought transmission process, the transmission direction, ending order, transmission type and drought occurrence range expansion trend of meteorological-vegetation drought are analyzed to clarify the meteorological-vegetation drought transmission process.

[0089] The characteristic variables of the meteorological drought to vegetation drought transmission process include the lead time, lag time, convergence amount, branch amount and area difference.

[0090] Specifically, the step includes the following contents,

[0091] 1. Lead time: the lead time is the difference between the starting time of the vegetation drought event and the starting time of the meteorological drought event, which represents the length of time for the transmission of meteorological drought to vegetation drought. Unit: month.

[0092] When the lead time is positive, it indicates that the meteorological-vegetation drought is positive transmission; otherwise, it indicates that the meteorological-vegetation drought event is negative transmission.

[0093] 2. Lag time: the lag time is the difference between the ending time of the vegetation drought event and the ending time of the meteorological drought event, which represents the length of time for the vegetation drought to extend relative to the meteorological drought. Unit: month.

[0094] When the difference between the ending time of the vegetation drought event and the ending time of the meteorological drought event is positive, it indicates that the vegetation drought ends later than the meteorological drought; otherwise, it indicates that the vegetation drought ends earlier than the meteorological drought.

[0095] 3. Convergence amount: the convergence amount is the number of meteorological drought events. Unit: number.

[0096] When the convergence amount is equal to 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is one-to-one or one-to-many; when the convergence amount is greater than 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is many-to-one or many-to-many.

[0097] 4. Branch quantity: the branch quantity is the number of vegetation drought events. Unit: pieces.

[0098] When the branch quantity is equal to 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is one-to-one or many-to-one; when the branch quantity is greater than 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is one-to-many or many-to-many.

[0099] 5. Area difference: the area difference is the difference between the projection areas of meteorological drought and vegetation drought on the geographic coordinate plane, indicating the change of the occurrence range of meteorological-vegetation drought. Unit: km 2 .

[0100] When the area difference is positive, it indicates that the occurrence range of meteorological-vegetation drought shows a shrinking trend; otherwise, it indicates that the occurrence range of meteorological-vegetation drought shows an expanding trend.

[0101] In this embodiment, a meteorological-vegetation drought transmission feature recognition system based on a three-dimensional clustering algorithm is also provided, which is used to implement the above-mentioned method. The system comprises,

[0102] 1. Drought event recognition module: based on a three-dimensional clustering algorithm, the drought patches are determined according to the drought index size of meteorological drought and vegetation drought, and all meteorological drought events and vegetation drought events in the research period are recognized according to the drought patch area and the size of the overlapping area of drought patches in adjacent months;

[0103] 2. Drought event analysis module: the dynamic changes of the identified meteorological drought events and vegetation drought events are analyzed from multiple drought event characteristic variables, and the characteristic variables of the corresponding drought events are determined;

[0104] 3. Drought event pairing: based on the drought transmission recognition rules of three-dimensional spatiotemporal clustering, the meteorological drought events and vegetation drought events are matched according to the overlapping time and overlapping area size of meteorological drought time and vegetation drought events in time and space dimensions, to obtain the meteorological-vegetation drought event pair;

[0105] 4. Event pair transmission type determination module: according to the number of meteorological drought events and the number of vegetation drought events, the drought transmission type of the meteorological-vegetation drought event pair is divided into four categories: one-to-one, one-to-many, many-to-one and many-to-many;

[0106] 5. Drought transmission process analysis module: based on the characteristic variables of the transmission process from meteorological drought to vegetation drought, the transmission direction, ending order, transmission type and drought occurrence range shrinking-expanding trend of meteorological-vegetation drought are analyzed to clarify the transmission process of meteorological-vegetation drought.

[0107] Embodiment two

[0108] In this embodiment, taking the meteorological drought and vegetation drought of North China Plain from 1982 to 2020 as an example, the execution process of the meteorological-vegetation drought transmission characteristic identification method based on the three-dimensional clustering algorithm is specifically explained.

[0109] North China Plain, also known as Huang-Huai-Hai Plain, is the second largest plain in China, a typical alluvial plain, belonging to semi-humid temperate continental monsoon climate, located between 31 ° N-43 ° N, 110 ° E-122 ° E, with an area of about 540,000 km 2 , and an average elevation of 50 m below. In recent years, the degree of vegetation drought has intensified, and the problems of ecological vulnerability and water resource shortage are prominent, which has greatly affected the sustainable development of the economy and society of the region. Due to the spatial heterogeneity of drought, the drought conditions in different sub-regions may differ. Based on the administrative division of the study area, North China Plain is divided into 5 sub-regions, namely Beijing, Tianjin, Shandong, Hebei, and Henan.

[0110] In this embodiment, the monthly VHI remote sensing data set from 1982 to 2020 provided by the STAR product of the National Oceanic and Atmospheric Administration (NOAA) of the United States (https: / / www.star.nesdis.noaa.gov / ) is used, with a spatial resolution of 4 km. The data is preprocessed using ArcGIS software, and through resampling, cropping and operation, the monthly VHI spatial data set with a spatial resolution of 0.1 ° x 0.1 ° for North China Plain is obtained.

[0111] In this embodiment, the monthly SPEI spatial data set with a spatial resolution of 0.1 ° x 0.1 ° is calculated using the FLDAS data set (https: / / disc.gsfc.nasa.gov / datasets / ). The FLDAS data set integrates CHIRPS and MERRA-2 remote sensing satellite data information, and contains a variety of variables related to climate fields (such as evapotranspiration, soil moisture, heat flux, surface radiation and total rainfall rate). Under the driving of multiple meteorological data, the FLDAS data set is generated by a global-scale high-resolution land surface data assimilation system.

[0112] In this embodiment, the meteorological-vegetation drought event pair matching and quantitative identification of the transmission characteristics of the meteorological-vegetation drought event pair are carried out using the meteorological-vegetation drought transmission characteristic identification method based on the three-dimensional clustering algorithm.

[0113] 1. Meteorological drought event and vegetation drought event identification results

[0114] Using the three-dimensional clustering method, 56 meteorological drought events were identified in the North China Plain from 1982 to 2020. In the 1980s, 1990s, 2000s, and 2010s, the number of vegetation drought events was 9, 22, 40, and 17, respectively. The mean duration of drought was 2 months, 2.85 months, 2.25 months, and 2.16 months, respectively. The mean area affected by drought was 13.45×10 4 km 2 , 15.63×10 4 km 2 , 19.56×10 4 km 2 , and 16.38×10 4 km 2 , respectively. The mean intensity of drought was 3.28×10 4 ·month·km 2 , 13.51×10 4 ·month·km 2 , 14.44×10 4 ·month·km 2 , and 11.28×10 4 ·month·km 2 , respectively. The mean intensity of drought was 0.11, 0.26, 0.36, and 0.25, respectively.

[0115] From 1982 to 2020, 81 vegetation drought events were identified in the North China Plain. In the 1980s, 1990s, 2000s, and 2010s, the number of vegetation drought events was 32, 27, 15, and 7, respectively. The mean duration of drought was 3.03 months, 2.74 months, 2.27 months, and 2.29 months, respectively. The mean area affected by drought was 4.30×10 4 km 2 , 4.65×10 4 km 2 , 3.69×10 4 km 2 , and 1.99×10 4 km 2 , respectively. The mean intensity of drought was 1.49×10 4 ·month·km 2 , 1.12×10 4 ·month·km 2 , 0.83×10 4 ·month·km 2 , and 0.50×10 4 ·month·km 2The average intensity of drought is 0.10, 0.09, 0.10, 0.11, respectively.

[0116] Since the intensity of the vegetation drought event from January 1982 to June 1982 is the largest (No. V4), this vegetation drought event is selected as a typical drought event for analysis, and the dynamic evolution process is shown in FIG. 4. Figure 2 This drought event lasted for 6 months, and the average value of drought duration was 3.06 months. The larger value was located in the northwest of Shandong Province and the northeast of Henan Province. This drought event began to emerge in January 1982, then reached the peak in April 1982, at which time the drought affected area was the largest, and finally the drought degree gradually decreased from April 1982 to June 1982. It can be seen that the intensity and the affected area both showed an increasing trend first and then a decreasing trend, and the intensity and the affected area showed a relatively consistent change trend. Overall, the typical vegetation drought event No. V4 reflects the drought process of occurrence-strengthening-decay-extinction.

[0117] 2. Meteorological-vegetation drought event pair

[0118] From 1982 to 2020, 13 meteorological-vegetation drought event pairs were successfully matched in the North China Plain. Among these meteorological-vegetation drought event pairs, there were 7 pairs of type “one-to-one”, 4 pairs of type “one-to-many”, 1 pair of type “many-to-one”, and 1 pair of type “many-to-many”. In addition, the maximum area difference between the meteorological drought event and the vegetation drought event appeared in event pair No. C-12, with an area difference of 31.35×10 4 km 2 The meteorological drought event and the vegetation drought event in this meteorological-vegetation drought event pair were No. M47 and No. V78, respectively, with a lag time of 0 month, a convergence amount of 1, and a branching amount of 1. This event pair belongs to the type “one-to-one”. The second highest area difference appeared in event pair No. C-7, with an area difference of 28.81×10 4 km 2 The meteorological drought event was No. M15, and the vegetation drought events were No. V54 and No. V55. The lag time was 0 month, the convergence amount was 1, and the branching amount was 2. This event pair belongs to the type “one-to-many”. In the type “many-to-one” (event pair No. C-13), the area difference between the meteorological drought event and the vegetation drought event was 21.74×10 4 km 2, the meteorological drought event is numbered as M48, M49, the vegetation drought event is numbered as V79, the lag time is 0 month, and the convergence quantity and the branch quantity are 2 and 1 respectively. In the ''many-to-many'' type (event pair number C-8), the meteorological drought event is numbered as M16, M17, M18, the vegetation drought event is numbered as V57, V58, V59, the area difference between the meteorological drought event and the vegetation drought event is 22.75*10 4 km 2 , the lag time is 2 and -1 month respectively, and the convergence quantity and the branch quantity are both 3.

[0119] Table 1 Characteristic variables of meteorological-vegetation drought event pair transmission process

[0120]

[0121] In the 13 meteorological-vegetation drought event pairs, the maximum value (7 months) of the meteorological drought duration appears in the C-8 event pair, which belongs to the ''many-to-many'' type, and the vegetation drought duration of the event pair is 4 months; and the maximum value (9 months) of the vegetation drought duration appears in the C-2 event pair, which belongs to the ''one-to-one'' type, and the meteorological drought duration of the event pair is 2 months. Meanwhile, the maximum value (39.93*10 4 km 2 ) of the meteorological drought influence area appears in the C-7 event pair, which belongs to the ''one-to-many'' type, the vegetation drought influence area of the event pair is 11.12*10 4 km 2 , and the meteorological-vegetation drought area difference is 28.81*10 4 km 2 ; and the maximum value (13.68*10 4 km 2 ) of the vegetation drought influence area appears in the C-2 event pair, which belongs to the ''one-to-one'' type, the meteorological drought influence area of the event pair is 4.14*10 4 km 2 , and the meteorological-vegetation drought area difference is -9.54*10 4 km 2 . The typical meteorological-vegetation drought event pairs of ''one-to-one'', ''one-to-many'', ''many-to-one'' and ''many-to-many'' are screened out respectively, as shown in Figure 3 , Figure 4 , Figure 5 , Figure 6 . It can be seen that the meteorological-vegetation drought event transmission has a time lag effect to a certain extent, and the occurrence range of the vegetation drought is generally smaller than that of the meteorological drought.

[0122] By adopting the above technical scheme disclosed in the present application, the following beneficial effects are obtained:

[0123] The application provides a meteorological-vegetation drought transmission characteristic identification method and system based on a three-dimensional clustering algorithm, which can match the characteristics of drought events in the time and space dimensions, quantitatively identify the characteristic changes in the meteorological-vegetation drought transmission process, reveal the transmission characteristics of meteorological drought to vegetation drought, and clarify the transmission mechanism of meteorological drought to vegetation drought. The method and system can quantitatively identify the characteristic changes in the meteorological-vegetation drought transmission process, reveal the transmission characteristics of meteorological drought to vegetation drought, clarify the internal relationship, response characteristics and transformation mechanism between meteorological drought and vegetation drought, and provide scientific and technological support for accurately predicting and preventing vegetation drought.

[0124] The above merely describes the preferred embodiments of the application, and it should be noted that, for those skilled in the art, some improvements and refinements can be made without departing from the principles of the application, and these improvements and refinements should also be considered as the protection scope of the application.

Claims

1. A method for identifying meteorological-vegetation drought transmission features based on a three-dimensional clustering algorithm, characterized in that: Includes the following steps, S1. Drought Event Identification: Based on the three-dimensional clustering algorithm, drought patches are identified according to the drought index of meteorological drought and vegetation drought. Based on the area of ​​drought patches and the overlap area of ​​drought patches in adjacent months, all meteorological drought events and vegetation drought events during the study period are identified. Step S1 specifically includes the following: S11. Determine the drought index and threshold values ​​for meteorological drought and vegetation drought; the drought index for meteorological drought is the standardized precipitation evapotranspiration index, and the threshold value is -1; the drought index for vegetation drought is the vegetation health index, and the threshold value is 0.

4. S12. Select grids with drought indices below the corresponding threshold on a monthly scale to form drought patches; S13. Determine whether the area of ​​the drought patch is less than the patch area threshold. If so, the drought patch is considered invalid and discarded; otherwise, it is retained and proceeds to S14. S14. Calculate the overlap area between drought patches in adjacent months and determine whether the overlap area is less than the patch overlap area threshold. If so, mark the two drought patches as two independent drought events; otherwise, mark the two drought patches as the same drought event. S15. Identify and obtain all drought events within the research period, delete drought events whose duration is less than the duration threshold, and number the remaining drought events to obtain meteorological drought events and vegetation drought events. S2. Drought Event Analysis: The dynamic changes of the identified meteorological drought events and vegetation drought events were analyzed by analyzing multiple characteristic variables of drought events to determine the characteristic variables of the corresponding drought events; Step S2 specifically involves analyzing the dynamic changes of meteorological drought events and vegetation drought events from five characteristic variables of drought events: drought duration, affected area, intensity, strength, and centroid, in order to determine the characteristic variables of the corresponding drought events. S3, drought event pairing: Based on the three-dimensional spatiotemporal clustering-based drought transmission identification rules, meteorological drought events and vegetation drought events are matched according to the overlap time and overlap area of ​​meteorological drought events and vegetation drought events in the temporal and spatial dimensions to obtain meteorological-vegetation drought event pairs. Step S3 specifically includes the following: S31. Sort meteorological drought events and vegetation drought events in chronological order; S32. Determine whether there is an intersection between meteorological drought and vegetation drought events in time and space. If the overlap time is greater than the overlap time threshold and the overlap area is greater than the dimensional overlap area threshold, then the meteorological-vegetation event pair is considered valid; otherwise, it is not valid. S33. Number the successfully matched meteorological-vegetation event pairs in chronological order; S4. Event pair delivery type determination: Based on the number of meteorological drought events and the number of vegetation drought events, the drought transmission type of meteorological-vegetation drought event pairs is divided into four categories: one-to-one, one-to-many, many-to-one, and many-to-many. Step S4 specifically involves classifying the drought transmission type of the meteorological-vegetation drought event pair into four categories based on the number of meteorological drought events and the number of vegetation drought events, namely: One-to-one: A meteorological drought event triggers a vegetation drought event; One-to-many: A meteorological drought event triggers multiple vegetation drought events; Many to one: Multiple meteorological drought events trigger a single vegetation drought event; Many-to-many: Multiple meteorological drought events triggered multiple vegetation drought events; S5. Analysis of drought transmission processes: Based on the characteristic variables of the meteorological drought to vegetation drought transmission process, this study analyzes the transmission direction, termination sequence, transmission type, and the trend of expansion and contraction of the drought occurrence range in order to elucidate the meteorological drought to vegetation drought transmission process. Step S5 specifically includes the following: When the difference between the start times of vegetation drought events and meteorological drought events is positive, it indicates that the meteorological-vegetation drought is transmitted in a positive direction; conversely, it indicates that the meteorological-vegetation drought is transmitted in a negative direction. When the difference between the end times of vegetation drought and meteorological drought events is positive, it indicates that vegetation drought ends later than meteorological drought; conversely, it indicates that vegetation drought ends earlier than meteorological drought. When the number of meteorological drought events is equal to 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is one-to-one or one-to-many; when the number of meteorological drought events is greater than 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is many-to-one or many-to-many. When the number of vegetation drought events is equal to 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is one-to-one or many-to-one; when the number of vegetation drought events is greater than 1, it indicates that the transmission type of the meteorological-vegetation drought event pair is one-to-many or many-to-many. When the difference between the projected areas of meteorological drought and vegetation drought on the geographic coordinate plane is positive, it indicates that the range of meteorological-vegetation drought is shrinking; conversely, it indicates that the range of meteorological-vegetation drought is expanding.

2. The meteorological-vegetation drought transmission feature identification method based on three-dimensional clustering algorithm according to claim 1, characterized in that: The overlap time threshold is 0.

3. A meteorological-vegetation drought transmission feature identification system based on a three-dimensional clustering algorithm, characterized in that: The system is used to implement the method described in any one of claims 1 to 2, and the system comprises: Drought event identification module: Based on a three-dimensional clustering algorithm, drought patches are identified according to the drought index of meteorological drought and vegetation drought. Based on the area of ​​drought patches and the overlap area of ​​drought patches in adjacent months, all meteorological drought events and vegetation drought events during the study period are identified. Drought event analysis module: Analyzes the dynamic changes of identified meteorological drought events and vegetation drought events from multiple drought event characteristic variables to determine the characteristic variables of the corresponding drought events; Drought event pairing: Based on the drought transmission identification rules of three-dimensional spatiotemporal clustering, meteorological drought events and vegetation drought events are matched according to the overlap time and overlap area in the temporal and spatial dimensions to obtain meteorological-vegetation drought event pairs; Event Pair Transmission Type Determination Module: Based on the number of meteorological drought events and the number of vegetation drought events, the drought transmission type of meteorological-vegetation drought event pairs is divided into four categories: one-to-one, one-to-many, many-to-one, and many-to-many. The drought transmission process analysis module analyzes the transmission direction, termination sequence, transmission type, and expansion / contraction trend of the drought occurrence range based on the characteristic variables of the meteorological drought to vegetation drought transmission process, so as to clarify the meteorological drought transmission process.

Images