A drought event identification method, system, computer device and storage medium

Abstract

The application provides a drought event identification method, system, computer device and storage medium, and belongs to the field of drought identification, and comprises the following steps: obtaining original SAR data in a continuous time period of a research area; obtaining standardized monthly scale 200m resolution deformation data of the research area based on the original SAR data; decomposing the monthly scale 200m resolution deformation data and retaining residual term data reflecting abnormal hydrological changes; calculating IDSI values of corresponding time scales according to the residual term data of different time scales to form a multi-scale drought index sequence IDSI; and identifying drought events according to the multi-scale drought index sequence IDSI. The application does not need to rely on traditional observation data, improves the spatial coverage integrity and real-time performance of drought evaluation through high spatial resolution and continuous time sequence monitoring, realizes accurate description of hydrological abnormalities with the help of the correlation between surface deformation and deep hydrological processes, and the multi-scale index sequence can adapt to the time scale demand of different drought types, so that the comprehensiveness, accuracy and applicability of drought identification are significantly improved.

Description

Technical Field

[0001] This invention belongs to the field of drought identification technology, specifically relating to a drought event identification method, system, computer equipment, and storage medium. Background Technology

[0002] Drought is a common meteorological disaster worldwide, characterized by its wide coverage, long duration, high frequency, and broad impact. With the increasing trend of global warming, drought events are becoming more frequent. Unlike sudden disasters such as floods and typhoons, droughts often develop slowly, but once formed, they severely impact agricultural production, water supply, ecosystem stability, and socio-economic structures. Therefore, drought is considered one of the most far-reaching and destructive disasters affecting human society. Analyzing the evolution of drought can help improve water resource management efficiency, improve the ecological environment, and provide important support for climate change research and disaster prevention and mitigation.

[0003] Drought manifests in various forms, including meteorological drought, agricultural drought, hydrological drought, and socioeconomic drought. Drought is a complex phenomenon caused by multiple factors, and its formation is often determined by indices to quantify its severity. Different drought indices consider significantly different influencing factors and employ different calculation principles. Furthermore, the changing characteristics of drought are complex and variable, making it difficult to define a universally applicable drought index. Consequently, there is still no widely accepted, unified standard for defining a drought index. Classic indices for measuring drought include the Standardized Precipitation Index (SPI), which calculates an index by fitting precipitation data to the data distribution; the Palmer Drought Severity Index (PDSI), which assesses soil moisture in a region by analyzing temperature and precipitation data, thereby evaluating the degree of drought; the Standardized Precipitation Evapotranspiration Index (SPEI), which combines the advantages of SPI and PDSI, considering not only precipitation but also temperature and evapotranspiration, and can accurately reflect the impact of climate change on drought; the Standardized Runoff Index (SRI), similar to SPI, assesses the anomalies of watershed runoff by standardizing daily, monthly, or annual runoff data; and the Standardized Groundwater Index (SGI), which compares actual observations with long-term historical data and normalizes them using standard deviation to assess the state of groundwater resources. However, these indices mostly rely on traditional meteorological and hydrological observation data, which have limitations in terms of spatial coverage density, real-time performance, or reflection of deep hydrological processes.

[0004] With the development of Earth observation technology, novel monitoring methods based on satellite remote sensing have provided a completely new perspective for drought research. The core observation target of GRACE (Gravity Recovery and Climate Experiment) and its follow-up mission GRACE-FO (GRACE Follow-On) is Terrestrial Water Storage (TWS), which integrates components such as surface water, soil water, and groundwater. Its observation values ​​directly reflect changes in water storage caused by drought or flooding, and it has wide applications in drought research. Using GRACE satellite observation data, Yirdaw et al. developed the Total Storage Deficit Index (TSDI) system, which revealed the hydrological evolution characteristics of the Canadian grassland drought in 2002–2003; Sinha et al. integrated surface and groundwater storage information to create and apply the Water Storage Deficit Index (WSDI), revealing the spatiotemporal evolution characteristics of drought in India; Zhao et al. developed a global grid drought severity index, the Drought Severity Index (GRACE-DSI), which performed well and provided a new method for global, large-scale drought monitoring; Thomas et al. proposed the GRACE Groundwater Drought Index (GGDI) to study groundwater drought in California, USA; Zhong et al. proposed a daily standardized GRACE reconstructed water storage anomaly index applied to nine outflow river basins in China, providing valuable reference for drought assessment in various regions. GNSS (Global Navigation Satellite System) networks can monitor deformation caused by changes in surface loads, thus indirectly reflecting changes in water storage, making it possible to study drought using GNSS data. Ferreira et al. proposed a new drought index (Drought Index based on GPS-observed Vertical Crustal Deformations, DIVCD) using GNSS-monitored vertical crustal deformation, which can effectively monitor drought conditions in Brazil. By using vertical crustal deformation data collected from GNSS networks and combining it with the Slepian basis function method to infer changes in surface water height, Jiang et al. proposed a drought monitoring tool based on GNSS data (GNSS-DSI). Tang et al. proposed a new multi-scale GNSS drought index (MGDI), which combines GNSS data and changes in water storage, and can more comprehensively describe hydrological drought.

[0005] In summary, while existing drought monitoring technologies based on GRACE satellites can achieve continuous observation on a global scale, their spatial resolution is limited, making it difficult to accurately depict the fine spatiotemporal characteristics of regional hydrological drought. Although GNSS technology has high spatiotemporal resolution, it is limited by the sparse distribution of ground station networks and can only be used for monitoring local areas. Neither technology can achieve high-resolution, large-scale continuous drought monitoring, nor can they accurately capture hydrological anomalies caused by deep hydrological processes, thus failing to meet the actual needs of regional refined drought monitoring and identification. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a drought event identification method. This method regards the high spatiotemporal resolution surface deformation sequence obtained by InSAR as a physical response indicator of hydrological drought, providing a new perspective and method for drought monitoring research.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A method for identifying drought events includes the following steps: Acquire raw SAR data for the study area over a continuous time period; The original SAR data is processed using time-series InSAR to retrieve surface deformation data of the study area. Discrete high-coherence points are selected from the surface deformation data, and the corresponding surface deformation data are extracted. The study area is divided into a 200m×200m regular grid to form multiple 200m grid pixels covering the entire area. The surface deformation data is then uniformly converted into continuous deformation data with a spatial resolution of 200m. Based on continuous deformation data with a spatial resolution of 200m, and using the natural month as the time unit, the average surface deformation of each 200m grid cell in the study area at all observation times in each month is calculated to obtain standardized monthly-scale 200m resolution deformation data of the study area. The study area was divided into 200m×200m regular grids to form multiple 200m grid pixels covering the entire area; the standardized monthly 200m resolution deformation data of the study area was decomposed to eliminate trend and seasonal terms, and retain residual terms that reflect abnormal hydrological changes; the IDSI values ​​of the corresponding time scales were calculated based on the residual terms data of different time scales to form a multi-scale drought index sequence IDSI. Drought events are identified based on the multi-scale drought index sequence IDSI.

[0008] Preferably, the StaMPS time-series InSAR analysis method is used to perform time-series analysis on the original SAR data to retrieve the surface deformation data of the study area from the original SAR data.

[0009] Preferably, the STL decomposition method is used to decompose and eliminate the trend and seasonal terms in the standardized monthly 200m resolution deformation data of the study area, while retaining the residual terms that reflect abnormal hydrological changes.

[0010] Preferably, the step of calculating the IDSI value for the corresponding time scale based on the residual data at different time scales to form a multi-scale drought index sequence IDSI specifically involves: For different time scales, a sliding window method is used to accumulate data and calculate the IDSI value for the corresponding time scale to form a multi-scale drought index sequence IDSI; ; in, i Representative year, j Representing the month, R ij yes i Year j The residual term obtained after removing the monthly trend and seasonal terms. It is all years j The average value of lunar deformation, σ j It is all years j The standard deviation of the lunar deformation.

[0011] Preferably, the identification of drought events based on the multi-scale drought index sequence (IDSI) specifically includes the following steps: The multiscale drought index sequence IDSI is labeled as X1, X2, X3, ..., X m , where m is the number of data points; For each pair of data points in the multi-scale drought index series IDSI, the statistic is calculated using a sign function. S ; Calculate the statistic by counting the number of samples with the same value in a statistical sequence. S The variance Var(S); based on the statistic S Given the variance Var(S), calculate the standardized Z-statistic; The calculated p-value is compared with a significance level of 0.05. A p-value < 0.05 indicates a significant trend, while a p-value ≥ 0.05 indicates no significant trend. Z > 0 indicates a wetting trend, and Z < 0 indicates a drying trend. When p < 0.05 and Z < 0, it indicates a significant drying trend in the study area. If p < 0.05 but Z > 0, it indicates a significant wetting trend in the study area. The significance level is used to characterize the drought or wetting in the study area as a real, long-term, regular change.

[0012] Preferably, the original SAR data is Sentinel-1A ascending orbit data, covering the period from July 2015 to December 2023, including 235 scenes in 40 orbits and 228 scenes in 142 orbits, in the C-band, with VV polarization, and azimuth and range resolutions of 20m and 5m, respectively.

[0013] Preferably, the method further includes verifying the reliability of the multi-scale drought index sequence IDSI by using the standardized precipitation evapotranspiration index (SPEI) at various time scales of the stations in the study area; obtaining the standardized precipitation evapotranspiration index (SPEI) at various time scales of the stations in the study area includes: Meteorological station data from multiple stations in the study area were collected over a continuous period of time. The meteorological station data included average temperature, average wind speed at a height of 2m, solar radiation, average relative humidity, average air pressure, and average precipitation. The standardized precipitation evapotranspiration index (SPEI) for each time scale of the stations in the study area was calculated using the meteorological station data.

[0014] The present invention also provides a drought event identification system, comprising: The data acquisition module is used to acquire raw SAR data of the study area over a continuous time period. The data conversion module is used to perform time-series InSAR processing on the original SAR data, inverting the surface deformation data of the study area from the original SAR data; filtering out discrete high-coherence points from the surface deformation data and extracting the surface deformation data corresponding to the high-coherence points; dividing the study area into a 200m×200m regular grid to form multiple 200m grid pixels covering the entire area; uniformly converting the surface deformation data into continuous deformation data with a spatial resolution of 200m; based on the continuous deformation data with a spatial resolution of 200m, using the natural month as the time unit, calculating the average surface deformation of each 200m grid pixel in the study area at all observation times in each month, to obtain standardized monthly-scale 200m resolution deformation data of the study area. The data decomposition module is used to decompose the standardized monthly 200m resolution deformation data of the study area, eliminate the trend and seasonal terms in the data, and retain the residual terms that reflect abnormal hydrological changes; and calculate the IDSI values ​​of the corresponding time scales based on the residual terms data of different time scales to form a multi-scale drought index sequence IDSI. The drought identification module is used to identify drought events based on the multi-scale drought index sequence (IDSI).

[0015] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement any of the steps in the drought event identification method.

[0016] The present invention also provides a computer-readable storage medium storing a computer program that, when loaded by a processor, can execute any of the steps in the drought event identification method.

[0017] The drought event identification method provided by this invention has the following beneficial effects: This invention first acquires continuous time-series raw SAR data, leveraging its wide coverage to overcome the limitations of local GNSS monitoring. Then, it uses time-series InSAR processing to retrieve surface deformation data, selecting high-coherence points and combining this with a 200m×200m regular grid to convert discrete data into continuous 200m high-resolution deformation data across the entire domain. This completely compensates for the insufficient spatial resolution of GRACE, achieving a balance between large-scale and high-precision monitoring. Subsequently, the average deformation value of each grid cell is calculated on a monthly basis to obtain standardized monthly-scale deformation data, laying the foundation for accurate extraction of hydrological signals. Through decomposition processing, trend and seasonal interferences are eliminated, retaining only residual terms reflecting anomalous hydrological changes. Combined with multi-year climatological standardization, multi-scale IDSI is calculated to accurately capture anomalies caused by deep hydrological processes, solving the problem of traditional techniques' inability to characterize groundwater dynamics. Finally, drought events are identified based on multi-scale IDSI, adapting to the time scale requirements of different drought types. This invention not only achieves high-resolution, large-scale continuous drought monitoring, but also improves the accuracy and comprehensiveness of drought identification, fully meeting the actual needs of regional refined drought monitoring and water resource management. Attached Figure Description

[0018] To more clearly illustrate the embodiments and design schemes of the present invention, the accompanying drawings required for this embodiment will be briefly described below. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of the drought event identification method according to Embodiment 1 of the present invention; Figure 2 Spatial distribution of wet and dry trends at different scales; Figure 3 Correlation between IDSI and SPEI site data and SPEI product data; Figure 4 This presents the correlation statistics between IDSI and SPEI site data; among which, Figure 4 (a) is the statistical result on a 6-month scale. Figure 4 (b) is the statistical result on a 12-month scale; Figure 5Comparison of IDSI and SPEI at different sites on 6-month and 12-month timescales; Figure 6 Drought events identified by IDSI; Figure 7 The IDSI and SPEI lag times at different sites; among them, Figure 7 (a) is the 6-month lag time. Figure 7 (b) is the 12-month lag time; Figure 8 The correlation between IDSI and SPEI changes for site 53781 at different lag times. Detailed Implementation

[0020] To enable those skilled in the art to better understand and implement the technical solutions of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0021] Drought is a natural disaster affecting regional water resource security and ecosystem stability. Traditional drought indices still have limitations in spatial resolution and characterization of groundwater processes. Therefore, this invention proposes an InSAR Drought Severity Index (IDSI) based on InSAR time-series surface deformation data and applies it to drought monitoring and propagation analysis in the Beijing-Tianjin-Hebei Plain from 2015 to 2023. STL decomposition is used to eliminate long-term trends and seasonal terms in the deformation sequence. The IDSI is constructed using standardized residuals, and drought trend characteristics are analyzed using the Mann-Kendall test. The results show that the IDSI can effectively reveal the spatiotemporal pattern of regional hydrological drought, exhibiting higher stability and stronger drought indication capabilities at long-term scales (6 months and 12 months). The southern and central regions of the study area show a significant drying trend, while the northern region generally shows wetting. The major drought events identified by the IDSI are highly consistent with historical records, verifying its reliability in actual drought monitoring. A comparison with the meteorological drought index SPEI revealed a strong correlation between the two on 6-month and 12-month timescales, with the IDSI exhibiting a certain lag across multiple sites. Further lag correlation analysis showed that the average propagation time from meteorological drought to hydrological drought was approximately 4 months on a 6-month timescale and approximately 5 months on a 12-month timescale. This study demonstrates the significant advantages of the IDSI in characterizing groundwater-driven hydrological drought, providing new technical means and scientific basis for drought monitoring and regional water resource management.

[0022] Example 1 This embodiment selects the Beijing-Tianjin-Hebei Plain as the study area and proposes a method for obtaining the InSAR Drought Severity Index (IDSI) based on InSAR deformation time series. Using Sentinel-1A data from 2015 to 2023, the spatiotemporal distribution characteristics of drought in the Beijing-Tianjin-Hebei Plain at multiple time scales were obtained. The reliability of the IDSI results was verified using meteorological station SPEI and CHM_Drought SPEI products. Furthermore, the study on drought event identification and the propagation law from meteorological drought to hydrological drought in the study area was carried out.

[0023] Based on this, the present invention provides a drought event identification method, specifically an IDSI-based drought event identification method, such as... Figure 1 As shown, the method includes the following steps: Step 1: Collect multi-source basic data of the study area and preprocess the collected data.

[0024] Specifically, the study area in this embodiment is the Beijing-Tianjin-Hebei Plain. As an important agricultural region, the Beijing-Tianjin-Hebei Plain mainly includes parts of the plains of Beijing, Tianjin, and Hebei. The climate of the study area is a typical temperate monsoon climate, characterized by hot and rainy summers and cold and dry winters, with precipitation concentrated mainly in summer. The region has flat terrain and abundant water resources, but due to long-term high agricultural irrigation demands and uneven seasonal distribution of precipitation, water resources face certain pressures, especially during the dry season when the supply and demand contradiction becomes increasingly prominent. In recent years, the aridification trend in the North China Plain has intensified, with its frequency, duration, and severity continuously increasing. Multiple studies have revealed significant land subsidence caused by groundwater over-extraction in the Beijing-Tianjin-Hebei Plain, with an average subsidence rate reaching tens of millimeters per year, and the maximum subsidence in local groundwater funnel areas reaching approximately 120 mm / year. Its spatiotemporal evolution is highly consistent with changes in groundwater storage.

[0025] The multi-source basic data includes the original SAR data of the study area, meteorological station data, and standardized precipitation evapotranspiration index (SPEI).

[0026] Specifically, the original SAR data used in this invention is Sentinel-1A ascending orbit data, covering the period from July 2015 to December 2023, including two orbits, 40 and 142. Specifically, it includes 235 scenes of data in orbit 40 (i.e., 235 effective SAR remote sensing images of the Beijing-Tianjin-Hebei Plain taken by the Sentinel-1A satellite in ascending orbit numbered 40 between 2015 and 2023) and 228 scenes of data in orbit 142. The data is in the C-band, with VV polarization, and azimuth and range resolutions of 20m and 5m, respectively. As the basic data source for time-series InSAR processing, it can accurately capture surface deformation information.

[0027] Regarding meteorological data, this invention collected data on six elements from 129 meteorological stations in China from 2003 to 2023, including average temperature, average wind speed at 2m altitude, solar radiation, average relative humidity, average air pressure, and average precipitation, to calculate the standardized precipitation evapotranspiration index (SPEI) for each time scale at the stations in the study area.

[0028] The standardized precipitation evapotranspiration index (SPEI) used in this invention is derived from the China Hydrometeorological Dataset Drought (CHM_Drought) published by Zhang et al. This dataset is based on observational data from 2419 stations of the China Meteorological Administration, and was generated using angular distance weighting interpolation. It has a spatial resolution of 0.1° and spans from 1961 to 2022, encompassing multiple time scales including 2 weeks and 1-12 months. The SPEI serves as the benchmark for subsequent validation.

[0029] The collected data is then initially screened to remove outliers and severely missing records, ensuring data integrity.

[0030] Step 2: Process the raw SAR data after initial screening.

[0031] Step 2.1: Perform time-series InSAR processing on the raw SAR data using the StaMPS method.

[0032] Specifically, the StaMPS time-series InSAR analysis method is used to perform time-series analysis on the raw SAR data collected in step 1. This method identifies persistent and distributed scattering points in the study area and inverts the surface deformation data of a large study area from the raw SAR data, which is particularly suitable for monitoring complex areas.

[0033] Step 2.2: Extract discrete, highly coherent point surface shape data.

[0034] High coherence points were selected from the surface deformation data obtained by StaMPS time-series InSAR processing, and their corresponding surface deformation data were extracted. This data only reflects the discrete locations of high coherence points, without a uniform spatial resolution, and is non-continuous point-like deformation data. Among them, high coherence points are surface target points in synthetic aperture radar interferometry (InSAR) that maintain strong radar echo signal coherence and stable scattering characteristics in SAR images of different time phases. They are also often called permanent scatterers (PS points) and are the core data foundation for time-series InSAR inversion of surface deformation. At the same time, the study area was divided into a regular 200m×200m grid to form multiple 200m grid pixels covering the entire area.

[0035] Based on the effective stress principle and one-dimensional consolidation theory, the hydraulic mechanism in this invention is explained. When groundwater in an aquifer is extracted, the pore water pressure p decreases, while the total stress σ generated by the overlying rock and soil remains unchanged. Therefore, the effective stress σ′ increases accordingly. The relationship among the three is as follows:

[0036] (1) Total stress σ is the stress generated by the total weight of the overlying soil and rock mass, and in a stable stratum, it can usually be assumed to be constant. Effective stress σ' is the stress borne and transmitted by the soil or rock skeleton, and it directly controls the deformation (compression or expansion) of the particle skeleton. Pore water pressure p is the pressure exerted by the water within the rock pores.

[0037] Under arid conditions, groundwater is extensively extracted to compensate for insufficient surface water resources, causing a sharp drop in aquifer water levels and a significant decrease in pore water pressure p. Since the total stress σ remains essentially constant, according to formula (1), the decrease in pore water pressure will lead to an equal increase in effective stress σ′. This stress transfer shifts the load originally borne by pore water to the soil skeleton, thereby causing a decrease in the compaction and porosity of the skeleton structure. The compaction of the entire aquifer system macroscopically manifests as large-scale ground subsidence. The surface deformation data obtained by InSAR technology in this invention is a direct reflection of the consolidation and compression process of the aquifer system on the surface.

[0038] Step 2.3: Spatial resampling generates continuous deformation data with a resolution of 200m.

[0039] Since the high coherence points selected by temporal InSAR are discretely distributed, in order to facilitate subsequent calculations, the surface deformation data obtained in step 2.2 are uniformly converted into continuous deformation data with a spatial resolution of 200m, so that the entire study area forms a regular spatial grid, and each grid cell corresponds to a unique surface deformation value, making the data format uniform, facilitating subsequent spatial analysis and calculation, and laying the foundation for the construction of the drought index.

[0040] Step 2.4: Calculate the mean deformation of each 200m pixel by month.

[0041] Based on continuous deformation data with a spatial resolution of 200m, and using the natural month as the time unit, the average surface deformation of each 200m grid cell in the study area is calculated at all observation times within each month, thereby achieving time-scale normalization of the deformation data.

[0042] After calculating the monthly average, standardized monthly 200m resolution deformation data for the study area were obtained, namely the InSAR surface deformation time series. This data combines regular 200m spatial resolution and a regular monthly time scale, eliminating the problem of inconsistent time in the original observation data and providing standard data for subsequent trend and seasonal term elimination.

[0043] Step 3: STL decomposition eliminates trend and seasonal terms in the standardized monthly 200m resolution deformation data of the study area.

[0044] Since the InSAR surface deformation time series is simultaneously affected by multiple sources such as continuous consumption of groundwater resources, irrigation cycle and load elastic response, this invention first removes its long-term trend and seasonal terms, retains only the residual components that reflect abnormal hydrological changes, and further standardizes and constructs IDSI based on the climate state of the same month over many years.

[0045] Specifically, to eliminate the interference of long-term trends and seasonal factors on deformation data, this invention calculates the mean deformation by month and then uses the STL decomposition method to process the standardized monthly 200m resolution deformation data of the study area. The STL decomposition method decomposes the deformation time series into trend, seasonal, and residual terms through local weighted regression. Finally, the trend and seasonal terms are removed, and only the residual terms reflecting abnormal hydrological changes are retained, laying the foundation for the subsequent construction of the IDSI drought index.

[0046] Specifically, STL decomposition, based on the principle of local weighted regression, is used to fit the trend or seasonal pattern of local data, and standardizes the monthly 200m resolution deformation data of the study area. The analysis is decomposed into three additive components. These three terms are: a trend term (reflecting long-term changes, such as long-term subsidence caused by continuous groundwater depletion), a seasonal term (a fixed-period recurring pattern, such as periodic deformation caused by agricultural irrigation cycles), and a residual term (the remaining part after removing the trend and seasonal components; reflecting only deformation signals caused by anomalous hydrological changes). Interfering trend and seasonal terms are removed, and only the residual term, which characterizes drought-related anomalies, is retained for subsequent analysis. This residual term effectively reflects surface deformation caused by anomalous hydrological changes.

[0047] Specifically, the three components of the sum are represented as follows: (2) Here, Trend (T) represents long-term changes in the data, such as continuous growth, decline, or stagnation. Seasonal (S) represents recurring patterns within a fixed period (such as regular changes annually, quarterly, or monthly). Remainder (R) represents the components remaining after removing the trend and seasonal components.

[0048] Step 4: IDSI Index Calculation Specifically, based on the residual terms obtained from STL decomposition, the IDSI is constructed by standardization using the climatology of the same month over many years as a benchmark. According to formula (3), for different time scales such as 1, 3, 6, and 12 months, the IDSI values ​​of the corresponding time scales are calculated by using the sliding window cumulative data method, forming a multi-scale drought index sequence IDSI.

[0049] (3) in, i Representative years (2015-2023). j Representative months (January to December) R ij yes i Year j The residual term obtained after removing the monthly trend and seasonal terms. It is all years j The average value of lunar deformation, σ j It is all years j The standard deviation of the lunar deformation is calculated using a sliding window to accumulate data at different scales.

[0050] Step 5: Mann-Kendall test for drought trend To investigate the spatiotemporal distribution characteristics of drought in the Beijing-Tianjin-Hebei Plain, pixel-by-pixel trend analysis was performed on IDSI results at 1, 3, 6, and 12-month time scales using the Mann-Kendall nonparametric statistical test. First, given the IDSI time series X1, X2, X3, ..., X... m The sign difference of each pair of data points is calculated to obtain the statistic S. Then, the variance Var(S) is calculated by combining the possible duplicate values ​​in the sequence, and finally the standardized Z statistic is obtained. The calculated p value is compared with the significance level of 0.05. A significant trend is considered to exist when p < 0.05, where Z > 0 indicates a wetting trend and Z < 0 indicates a drying trend, thereby revealing the spatiotemporal evolution law of drought in the study area.

[0051] The inspection process can be summarized in the following steps: Given a time series X1, X2, X3, ..., X m , where m is the number of data points. The statistic is obtained by calculating the sign difference between each pair of data points in the time series. S :

[0052] (4) Where, sign(x) j -xi Let x be a sign function, if x j >x i Then sign(x) j -x i If x = 1; j <x i Then sign(x) j -x i If x = −1; j =x i Then sign(x) j -x i =0. To assess the significance of the trend, the MK test is based on... S The standardized Z-statistic is calculated using the following formula:

[0053] (5) The variance Var(S) is calculated to account for possible repeating values ​​in the sequence. (6) In the formula, g It is the number of knots. t p This represents the width of the p-th node (i.e., the number of samples with the same value). The calculated Z-value is compared with the standard normal distribution to obtain the corresponding p-value representing the probability of drought. When the p-value is less than the significance level (usually 0.05), the null hypothesis is rejected, indicating a significant trend in the data. When the p-value is greater than or equal to the significance level, no significant trend is considered. Z>0 indicates an upward trend, i.e., a wetting trend; Z<0 indicates a downward trend, i.e., a drying trend. When p<0.05 and Z<0, it indicates a significant drought trend in the study area; if p<0.05 but Z>0, it indicates a significant wetting trend. A significant drying trend indicates that the drying in this area is not a random phenomenon caused by short-term meteorological changes, but a real, long-term, regular change.

[0054] Pixel-by-pixel trend analysis results are as follows Figure 2As shown in the figure, the wet and dry changes in the study area exhibit a significant scale dependence. The 1-month timescale IDSI (IDSI-1) is significantly affected by monthly meteorological fluctuations, with almost no significant trend across the entire region. When the scale increases to IDSI-3, regional differentiation of wet and dry trends begins to appear: 13.9% of the area shows significant aridification, 13.2% shows significant humidification, and 72.9% shows no significant trend. Among these, Beijing and Langfang show a humidification trend, while southern Baoding, central Cangzhou, and western Handan begin to show aridification trends. When the timescale is 6 months, the trend significance intensifies: 24.9% shows significant aridification, 23.6% shows significant humidification, and 51.5% shows no significant trend. Among these, Beijing, Langfang, and western Tianjin still show a humidification trend, while southern Baoding, central Cangzhou, and northern Hengshui show significant continuous aridification areas. The annual-scale IDSI-12 shows the most significant trend, with 32.6% of the area showing significant aridification, 34.2% showing significant humidification, and 33.2% showing no significant trend. Among these, the humidification trend in areas such as Beijing, western Tianjin, Langfang, and northern Baoding has further intensified, forming large-scale continuous zones. Meanwhile, southern Baoding, central Cangzhou, and northern Hengshui have become the most significant continuous areas of long-term aridification. Overall, with increasing time scale, the aridification and humidification trends have gradually evolved from short-term, scattered changes to a pattern with clear regional consistency. The northern region has consistently been dominated by humidification, with a significant trend on medium- and long-term scales. Conversely, the central and southern parts of the study area, especially the strip-shaped area consisting of southern Baoding, central Cangzhou, and northern Hengshui, as well as western Hengshui, show a continuous aridification trend at all scales, indicating a higher long-term drought risk in these areas.

[0055] Step 6: IDSI Reliability Verification To verify the reliability of IDSI, this invention collected SPEI product data from the CHM_Drought dataset (hereinafter referred to as CHM_SPEI) and calculated the SPEI of each station (hereinafter referred to as SPEI) using meteorological station data, and compared and analyzed it with the IDSI estimated by this invention.

[0056] The reliability of the IDSI obtained in step 4 was verified by comparing it with the traditional meteorological drought index SPEI. On one hand, potential evapotranspiration was calculated using collected meteorological station data based on the Penman-Montes formula, and then the SPEI for each station at different time scales was calculated using water surplus and deficit. On the other hand, the SPEI product from the CHM_Drought dataset was directly used. The correlation between the IDSII time series and the CHM_SPEI and SPEI time series at 1, 3, 6, and 12-month time scales was calculated, and the characteristics of correlation changes at different scales were analyzed. The consistency at medium- and long-term scales such as 6 months and 12 months was particularly verified. Combined with correlation stability analysis of some stations, the ability of IDSI to identify drought characteristics was confirmed.

[0057] The core of calculating SPEI is to calculate the water surplus or deficit and then standardize it by fitting a probability distribution.

[0058] (7) In the formula D It's about the water balance. PE For precipitation, ET This represents potential evapotranspiration.

[0059] Potential evapotranspiration data needs to be calculated. The Penman-Montes formula is currently recognized internationally as one of the most accurate methods for calculating potential evapotranspiration. It is recommended and widely used by the Food and Agriculture Organization of the United Nations. This formula comprehensively considers the principles of energy balance and aerodynamics, and utilizes multiple meteorological factors such as temperature, radiation, humidity, and wind speed. It is an important tool in fields such as global drought monitoring, irrigation management, and climate change research. (8) In the formula, ET Potential evapotranspiration R n It's net radiation, 𝜌 a air density, C p It is the specific heat capacity of air. G This is soil heat flux (which is usually ignored). V PD It is the water vapor pressure difference. It is the specific humidity of dry and humid air. r s It is the resistance to transpiration in plants. It's air resistance.

[0060] After obtaining the potential evapotranspiration data, calculate the cumulative water surplus / deficit over a k-month timescale: (9) In the formula The cumulative water surplus or deficit over a k-month timescale.

[0061] SPEI is calculated using a Log-Logistic distribution, and the cumulative probability is calculated. P : (10) α , β , These are the location, scale, and shape parameters of the fitted distribution, respectively.

[0062] Finally, the cumulative probability values ​​obtained from the fitting are standardized to obtain the SPEI values ​​that follow a standard normal distribution.

[0063] (11) Φ −1 It is the inverse function of the standard normal distribution.

[0064] The final analysis results are as follows Figure 3 As shown in the figure, the time span for IDSI and SPEI is from June 2015 to December 2023, while the time span for CHM_SPEI is from June 2015 to December 2022. The results show that the correlation between IDSI and CHM-SPEI and SPEI varies with the time scale; the larger the time scale, the stronger the correlation. When the time scale is 1 month and 3 months, the correlation is low, with IDSI having a correlation of 0.19 with both at the 1-month scale and 0.24 at the 3-month scale. The correlation improves at the 6-month scale compared to the other two scales, with CHM_SPEI having a correlation of 0.38 with IDSI and SPEI having a correlation of 0.43, slightly higher than the correlation between CHM_SPEI and IDSI. At the 12-month scale, the correlation between IDSI and both reaches its highest point, at 0.54 and 0.58 respectively, with SPEI still having a slightly higher correlation with IDSI than CHM_SPEI with IDSI.

[0065] To further analyze the reliability of IDSI products, this invention calculated the correlation between SPEI data and IDSI data at 6-month and 12-month scales for each meteorological station, and represented the stations with four different colors according to the magnitude of the correlation. Figure 4 In the study, at a 6-month timescale, SPEI data from most meteorological stations showed a positive correlation with IDSI, with approximately 50% of stations exhibiting a correlation between 0 and 0.3, and about 22% of stations showing a correlation between IDSI and SPEI between 0.3 and 0.5. The highest correlation between SPEI and IDSI at this timescale was 0.48. At a 12-month timescale, the correlation improved, with about 30% of meteorological stations showing a correlation above 0.3, and the highest correlation reaching 0.67. This indicates that IDSI can better reflect climate and drought trends over longer timescales. Furthermore, the correlation between SPEI and IDSI data from the same station showed good stability across different timescales. The strength of the correlation and its spatial distribution pattern at each station did not change significantly with time scale changes, indicating a certain degree of consistency in the response of SPEI and IDSI to drought processes across different timescales. This further supports the reliability of IDSI in identifying drought characteristics. This invention selected two stations and compared SPEI and IDSI data at two different timescales. The station locations are shown in [the table / image / image]. Figure 4 In (a), the comparison results are as follows: Figure 5As shown, data from station 64,633 at different time scales showed high correlations, with IDSI and SPEI exhibiting good trend consistency, with correlations of 0.48 and 0.67 at the six-month and 12-month scales, respectively. However, at station 53,781, the correlation was weaker, with correlations of 0.07 and 0.06 at the six-month and 12-month scales, respectively. A significant lag between IDSI and SPEI can be observed, which may be related to the local geographical location and the time required for meteorological drought to spread to hydrological drought.

[0066] Step 7: Drought event identification and historical verification based on IDSI Since the IDSI of this invention is calculated based on deformation, its data fluctuation is slightly smaller compared to SPEI. Using a fixed value would affect the reliability of the study. This invention uses the percentile threshold method to identify drought events, defining a drought event as a period of two consecutive months or more where the IDSI is below the 30th percentile threshold. Based on the IDSI-1 index (DSI-1 represents the IDSI index on a one-month timescale), this invention identified a total of 7 drought events between July 2015 and December 2023. Figure 6 As shown.

[0067] To assess the reliability of IDSI in monitoring actual drought events, this invention uses historical drought records from sources such as the *China Flood and Drought Disaster Bulletin*, the *Hebei Provincial Climate Bulletin*, and relevant research literature. Drought events (time and extent) identified by IDSI are compared one by one with historical records to verify their consistency and clarify the effectiveness of IDSI in monitoring actual drought events. The *China Flood and Drought Disaster Prevention Bulletin 2018* mentions that spring precipitation in the northern winter wheat region was 30% lower than usual, resulting in spring drought. According to the *China Flood and Drought Disaster Prevention Bulletin 2019*, the average precipitation in the study area in June 2019 was only 33 mm, a 55% decrease compared to the same period in previous years, indicating a rapid development of drought. Effective rainfall occurred in August, alleviating the drought. The *Hebei Provincial Climate Bulletin* records that in April 2021, precipitation was low, and in May, precipitation was more than 50% below average, with drought emerging and continuing to develop. Rainfall eased in June. In the spring of 2022, most areas of the Beijing-Tianjin-Hebei region experienced low rainfall, resulting in spring and summer drought. These historical drought records are consistent with the results identified by IDSI, demonstrating the reliability of IDSI in capturing drought events and showcasing its potential as a reliable drought monitoring tool for regional water resource management and disaster prevention.

[0068] Table 1 Comparison of IDSI-identified drought events and records Step 8: Analysis of the patterns of drought propagation To further improve the drought identification system, the Pearson lag correlation method was used to analyze the propagation characteristics of meteorological drought to hydrological drought. For SPEI and IDSI data at the same time scale from various stations, the IDSI data were processed with lags ranging from 1 to 12 months. The correlations at different lag times were calculated, and the lag time corresponding to the highest correlation was taken as the propagation time. The proportion of stations with different lag time periods was statistically analyzed to determine the average propagation time of meteorological drought to hydrological drought at 6-month and 12-month scales, providing a more comprehensive scientific basis for drought early warning.

[0069] This invention employs the most commonly used linear drought propagation method, the Pearson correlation coefficient method, to explore the propagation relationship between meteorological drought and groundwater drought. Using SPEI and IDSI, with the same time scale, to represent meteorological drought and hydrological drought respectively, at different lag times (1-12 months), the lag time corresponding to the maximum correlation coefficient is considered the propagation time from meteorological drought to hydrological drought. The formula for calculating the Pearson correlation coefficient is as follows:

[0070] (12) In the formula, R The correlation coefficient is denoted as n; n is the total length of the time series. X and Y are the means of sequences X and Y, respectively.

[0071] Because hydrological drought responds with a lag to meteorological drought, this invention utilizes the Pearson lag correlation method to infer the drought propagation time between them. For SPEI and IDSI data at the same time scale at each station, calculations are performed using IDSI data with a lag of 1-12 months and SPEI data, respectively. The results are as follows: Figure 7As shown, the correlation between IDSI and SPEI data significantly improved after different time lags. At the 6-month scale, 39% of stations had a correlation above 0.3, with a maximum correlation of 0.6; at the 12-month scale, 51% had a correlation above 0.3, with a maximum correlation of 0.83. Stations with lag times of 0-4 months accounted for 54% at the 6-month scale and 50% at the 12-month scale, indicating that meteorological drought in these areas spreads rapidly to hydrological drought, reflecting changes in hydrological drought quickly. Stations with lag times of 9-12 months accounted for 28% at the 6-month scale and 35% at the 12-month scale. The hydrological system at these stations responded significantly to meteorological drought, requiring a longer transmission process before it appeared in IDSI. These phenomena are usually related to factors such as regional groundwater recharge cycles, soil moisture content, and irrigation, meaning that short-term precipitation deficits do not immediately translate into hydrological drought, but rather require accumulated water deficits over multiple seasons to be reflected. Therefore, the propagation time from meteorological drought to hydrological drought is longer in these regions. In the study area, the average propagation time from meteorological drought to hydrological drought is 4 months on a 6-month timescale, and 5 months on a 12-month timescale.

[0072] To visually demonstrate the changes in correlation with lag months and to study the propagation time of drought, this invention selected meteorological station 53781 and calculated its correlation with SPEI by lag of the IDSI time series by 1-12 months at 6-month and 12-month scales. Figure 8 It can be seen that without lag treatment, the correlation between IDSI and SPEI is -0.07 and -0.06 on the 6-month and 12-month timescales, respectively, indicating almost no correlation. As the lag time changes, the correlation between IDSI and SPEI gradually increases, reaching a peak of 0.47 and 0.69 when IDSI lags SPEI by 10 months on both the 6-month and 12-month timescales. This also indicates that the propagation time from meteorological drought to hydrological drought at this station is 10 months on both timescales.

[0073] This invention, based on Sentinel-1 time-series InSAR technology, achieves drought monitoring at a resolution of 1 km, effectively solving the problems of insufficient spatial resolution and limited ability to characterize groundwater dynamics in traditional drought indices. Compared to indices such as SPEI and SPI, which are mainly controlled by meteorological factors, IDSI more closely approximates the actual aquifer compaction and water storage changes, and is particularly suitable for hydrological drought analysis in areas with groundwater over-extraction.

[0074] This invention innovatively proposes a multi-timescale hydrological drought index (IDSI) based on InSAR temporal deformation data and has been successfully applied to drought monitoring and assessment in the North China Plain from 2015 to 2023. Studies show that IDSI can effectively reveal the spatiotemporal distribution characteristics of hydrological drought in this region, exhibiting high stability, especially at 6-month and 12-month timescales. Trend analysis reveals a significant aridification trend in the southern and central parts of the study area, while the northern region mainly shows a wetting trend, providing important reference for regional water resource management. Comparison and validation between IDSI and the traditional meteorological drought index SPEI show that IDSI has higher correlation and reliability at longer timescales (6 months and 12 months), reaching 0.48 and 0.67, respectively. Furthermore, the drought events identified by IDSI are highly consistent with actual drought records, indicating that the index can effectively capture the evolution of drought events. This invention further reveals a lag effect in the propagation of meteorological drought to hydrological drought using lag correlation analysis, with average propagation times of 4 months and 5 months at 6-month and 12-month timescales, respectively.

[0075] This invention addresses the limitations of traditional drought indices, which rely on meteorological and hydrological observation data, suffer from insufficient spatial coverage density, lack of real-time performance, and difficulty in reflecting deep hydrological processes. Through innovative breakthroughs using SAR data and InSAR technology, it first acquires continuous time-series raw SAR data and then processes it using InSAR to retrieve surface deformation data. Leveraging the all-weather, all-time observation advantages and wide coverage of SAR technology, it overcomes the insufficient spatial density of traditional observation data, while the time-series data ensures real-time monitoring. Next, high-coherence points are selected and converted into continuous deformation data with a 200m spatial resolution, significantly improving spatial coverage density and enabling refined grid monitoring. Subsequently, through monthly-scale standardization and decomposition of trend and seasonal terms, residual terms reflecting abnormal hydrological changes are retained. Since surface deformation is directly related to deep hydrological processes such as changes in groundwater storage and soil moisture, this processing can accurately capture deep hydrological anomalies that traditional indices cannot reach. Finally, multi-scale IDSI values ​​are calculated based on the residual terms to form a complete drought index sequence for drought identification. This invention does not rely on traditional observation data. It improves the spatial coverage integrity and real-time performance of drought assessment through high spatial resolution (200m grid) and continuous temporal monitoring. It also achieves accurate characterization of hydrological anomalies by leveraging the correlation between surface deformation and deep hydrological processes. At the same time, the multi-scale index sequence can adapt to the time scale requirements of different drought types, significantly improving the comprehensiveness, accuracy and applicability of drought identification, and providing more reliable technical support for drought monitoring.

[0076] Based on the same inventive concept, the present invention also provides a drought event identification system, comprising: The data acquisition module is used to acquire raw SAR data for the study area over a continuous time period.

[0077] The data conversion module performs time-series InSAR processing on the raw SAR data, inverting the surface deformation data of the study area from the raw SAR data; it filters out discrete high-coherence points from the surface deformation data and extracts the surface deformation data corresponding to the high-coherence points; it divides the study area into a 200m×200m regular grid to form multiple 200m grid pixels covering the entire area; it uniformly converts the surface deformation data into continuous deformation data with a spatial resolution of 200m; based on the continuous deformation data with a spatial resolution of 200m, it calculates the average surface deformation of each 200m grid pixel in the study area at all observation times within each month, using the natural month as the time unit, to obtain the standardized monthly-scale 200m resolution deformation data of the study area.

[0078] The data decomposition module is used to decompose the standardized monthly 200m resolution deformation data of the study area, eliminate the trend and seasonal terms in the data, and retain the residual terms that reflect abnormal hydrological changes. Based on the residual terms data of different time scales, the IDSI values ​​of the corresponding time scales are calculated to form a multi-scale drought index sequence IDSI.

[0079] The drought identification module is used to identify drought events based on the multi-scale drought index sequence (IDSI).

[0080] Each module in the aforementioned drought event identification system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0081] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps in the drought event identification method embodiment. Specific implementation methods can be found in the method embodiment, and will not be repeated here.

[0082] Furthermore, the present invention also provides a non-transitory computer-readable storage medium containing instructions on which a computer program is stored. For example, a memory containing instructions that can be executed by a processor of a computer device to perform the above-described method. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc. When the computer program is executed by the processor, it can implement the steps in the drought event identification method embodiments. Specific implementation methods can be found in the method embodiments, which will not be repeated here.

[0083] Those skilled in the art will understand that embodiments of the present invention can provide methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0084] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0085] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0086] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0087] It should be noted that the specific embodiments described above enable those skilled in the art to more fully understand the present invention, but do not limit the present invention in any way. Therefore, although the present invention has been described in detail in this specification and embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the present invention; and all technical solutions and improvements that do not depart from the spirit and scope of the present invention are covered within the protection scope of the present invention patent. No reference numerals in the claims should be construed as limiting the scope of the claims. Any simple variations or equivalent substitutions of technical solutions that can be readily obtained by those skilled in the art within the scope of the technology disclosed in the present invention are within the protection scope of the present invention.

Claims

1. A method for identifying drought events, characterized in that, Includes the following steps: Acquire raw SAR data for the study area over a continuous time period; The original SAR data is processed using time-series InSAR to retrieve surface deformation data of the study area. Discrete high coherence points were selected from the surface deformation data, and the surface deformation data corresponding to the high coherence points were extracted; the study area was divided into 200m×200m regular grids to form multiple 200m grid pixels covering the entire area. The surface deformation data were uniformly converted into continuous deformation data with a spatial resolution of 200m; Based on continuous deformation data with a spatial resolution of 200m, and using the natural month as the time unit, the average surface deformation of each 200m grid cell in the study area at all observation times in each month is calculated to obtain standardized monthly-scale 200m resolution deformation data of the study area. The standardized monthly 200m resolution deformation data of the study area were decomposed to eliminate trend and seasonal terms, and retain residual data that reflect abnormal hydrological changes. The IDSI value for the corresponding time scale is calculated based on the residual data at different time scales, forming a multi-scale drought index sequence IDSI. Drought events are identified based on the multi-scale drought index sequence IDSI.

2. The drought event identification method according to claim 1, characterized in that, The StaMPS time-series InSAR analysis method was used to perform time-series analysis on the original SAR data, and the surface deformation data of the study area was obtained by inversion from the original SAR data.

3. The drought event identification method according to claim 1, characterized in that, The STL decomposition method was used to decompose and eliminate the trend and seasonal terms in the standardized monthly 200m resolution deformation data of the study area, while retaining the residual terms that reflect abnormal hydrological changes.

4. The drought event identification method according to claim 1, characterized in that, The process of calculating the IDSI value for the corresponding time scale based on the residual data at different time scales to form a multi-scale drought index sequence IDSI is as follows: For different time scales, a sliding window method is used to accumulate data and calculate the IDSI value for the corresponding time scale to form a multi-scale drought index sequence IDSI; ； in, i Representative year, j Representing the month, R ij yes i Year j The residual term obtained after removing the monthly trend and seasonal terms. It is all years j The average value of lunar deformation, σ j It is all years j The standard deviation of the lunar deformation.

5. The drought event identification method according to claim 1, characterized in that, The identification of drought events based on the multi-scale drought index sequence (IDSI) specifically includes the following steps: The multiscale drought index sequence IDSI is labeled as X1, X2, X3, ..., X m , where m is the number of data points; For each pair of data points in the multi-scale drought index series IDSI, the statistic is calculated using a sign function. S ; Calculate the statistic by counting the number of samples with the same value in a statistical sequence. S The variance Var(S); based on the statistic S Given the variance Var(S), calculate the standardized Z-statistic; The calculated Z value is compared with the standard normal distribution to obtain the corresponding p value representing the probability of drought; The calculated p-value is compared with a significance level of 0.

05. A p-value < 0.05 indicates a significant trend, while a p-value ≥ 0.05 indicates no significant trend. Z > 0 indicates a wetting trend, and Z < 0 indicates a drying trend. When p < 0.05 and Z < 0, it indicates a significant drying trend in the study area. If p < 0.05 but Z > 0, it indicates a significant wetting trend in the study area. The significance level is used to characterize the drought or wetting in the study area as a real, long-term, regular change.

6. The drought event identification method according to claim 1, characterized in that, The original SAR data is Sentinel-1A upswing data, spanning from July 2015 to December 2023, including 235 scenes from orbit 40 and 228 scenes from orbit 142. It is located in the C-band, with VV polarization, and has azimuth and range resolutions of 20m and 5m, respectively.

7. The drought event identification method according to claim 1, characterized in that, It also includes verifying the reliability of the multi-scale drought index sequence IDSI by using the standardized precipitation evapotranspiration index SPEI at various time scales of the study area stations. Obtaining the standardized precipitation evapotranspiration index (SPEI) for each time scale at the stations in the study area includes: Meteorological station data from multiple stations in the study area were collected over a continuous period of time. The meteorological station data included average temperature, average wind speed at a height of 2m, solar radiation, average relative humidity, average air pressure, and average precipitation. The standardized precipitation evapotranspiration index (SPEI) for each time scale of the stations in the study area was calculated using the meteorological station data.

8. A drought event identification system, characterized in that, include: The data acquisition module is used to acquire raw SAR data of the study area over a continuous time period. The data conversion module is used to perform time-series InSAR processing on the original SAR data and retrieve the surface deformation data of the study area from the original SAR data. Discrete high coherence points were selected from the surface deformation data, and the surface deformation data corresponding to the high coherence points were extracted; the study area was divided into 200m×200m regular grids to form multiple 200m grid pixels covering the entire area. The surface deformation data were uniformly converted into continuous deformation data with a spatial resolution of 200m; Based on continuous deformation data with a spatial resolution of 200m, and using the natural month as the time unit, the average surface deformation of each 200m grid cell in the study area at all observation times in each month is calculated to obtain standardized monthly-scale 200m resolution deformation data of the study area. The data decomposition module is used to decompose the standardized monthly 200m resolution deformation data of the study area, eliminate the trend and seasonal terms in the data, and retain the residual terms that reflect abnormal hydrological changes. The IDSI value for the corresponding time scale is calculated based on the residual data at different time scales, forming a multi-scale drought index sequence IDSI. The drought identification module is used to identify drought events based on the multi-scale drought index sequence (IDSI).

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is loaded by the processor, it is able to perform the steps of the method according to any one of claims 1 to 7.

Images