Method for monitoring abrupt change disasters of drought and flood in region with balanced abrupt change rate and abrupt change degree

Abstract

The application discloses a regional drought-flood abrupt change disaster monitoring method with balanced abrupt change rate and abrupt change degree, which comprises the following steps: obtaining soil humidity and precipitation data, calculating a standardized index SSI, SPI, and performing sliding average to obtain SPIM and SSIM, and identifying grid drought-flood properties; determining a lead day b through lead-lag regression of regional mean values of the two, screening drought-flood transition grid points, and counting regional drought-flood transition events; taking the day with the most drought-flood transition grid points as a reference day, performing difference to obtain a change slope and calculating a multiple parameter a; determining a drought-flood transition event occurrence day, synthesizing a sequence before and after the occurrence day, calculating an abrupt change rate DFAAV and an abrupt change degree DFAAM; normalizing and summing the two to obtain a drought-flood abrupt change index DFAAI, and screening drought-flood abrupt change events. The method can significantly improve the accuracy and reliability of drought-flood abrupt change event monitoring.

Description

Technical Field

[0001] This invention relates to the field of meteorological disaster monitoring technology, specifically to a method for monitoring regional drought-flood transition disasters where the rate and degree of rapid change are balanced. Background Technology

[0002] Against the backdrop of global warming, the frequency, intensity, and spatial extent of abrupt drought-flood transitions have increased significantly. As a complex disaster characterized by a rapid shift from extreme drought to extreme flooding, abrupt drought-flood transitions pose a far greater threat to water security, food production, and socio-economic stability than single extreme events due to their suddenness, complexity, and high destructiveness. Their impact is particularly pronounced on crop yields, especially for crops with weak drought and flood resistance. Existing indices primarily rely on precipitation data for calculations, neglecting the assessment of soil moisture's influence on drought severity, leading to misjudgments of events. Therefore, developing monitoring methods for abrupt drought-flood transitions is of paramount importance.

[0003] Accurate identification of drought and flood processes is fundamental to monitoring abrupt shifts between drought and flood. The essence of abrupt shifts is a rapid hydrological-meteorological transition involving interactions across multiple timescales. One source of its complexity stems from the inherent differences in the physical properties of the drought and flood phases—drought manifests as a slowly changing soil moisture deficit caused by a sustained lack of precipitation, while floods are primarily caused by concentrated bursts of heavy rainfall, exhibiting significant suddenness. Previous studies have largely relied on precipitation data to monitor abrupt shifts, generally neglecting the prior soil moisture background and its synergistic effect with subsequent precipitation, thus failing to fully reveal the systemic mechanisms of the transition from hydrological drought to flood. Therefore, incorporating soil moisture data into the abrupt shift index can enhance the accuracy of event identification.

[0004] Given the importance of soil moisture in drought monitoring, existing technologies have gradually incorporated it into the identification of abrupt drought-flood transition events. For example, determining drought-flood status through soil moisture anomaly percentage or constructing coupled models of soil moisture and precipitation for drought-flood transition prediction have all increased the weight of soil moisture in event identification. However, most current inventions are still limited to single-factor identification based on soil moisture or precipitation, while methods involving two-factor coupling mainly focus on prediction models. There is still a lack of a multi-factor drought-flood transition identification method suitable for scientific research scenarios such as climate characteristic analysis, model evaluation, and mechanism diagnosis.

[0005] Current drought-flood transition indices still have limitations in intensity assessment. The degree and rate of transition, as important components of these indices, have been improved through various methods to enhance their accuracy. For example, some studies employ dynamic optimization methods to quantify the cumulative effect of differences in state mean and recovery rates within a sliding window, reflecting the actual intensity of asymmetric drought-flood transition events. Other techniques optimize traditional drought-flood transition indices by assigning differentiated weight coefficients to different events or by using conversion coefficients and fixed weights to regulate the impact of the rate of transition in the overall index. While these methods have improved the weighting of degree and rate of transition to some extent, achieving a sufficient balance between the two in the evaluation system remains difficult, and there is still room for optimization in weight allocation. Especially in practical applications, the importance of the rate of transition is often underestimated, leading to biases in disaster risk assessment. As the extreme nature of events increases, a higher rate of transition often indicates a more drastic shift between drought and flood states, which can rapidly break through the buffering capacity of hydrological systems, exacerbate the disaster-bearing pressure on infrastructure, physiological stress on crops, and the decline in ecosystem stability, thereby significantly amplifying the catastrophic nature of the event.

[0006] Furthermore, given the increasingly short duration and high frequency of rapid drought-flood transition events, the scientific selection of spatiotemporal scales for their identification and judgment is becoming increasingly crucial. For short-cycle rapid drought-flood transition events, existing indices mostly use regional averaging methods for calculation, which have limitations in regional delineation due to excessively large scope or strong subjectivity, leading to misjudgments because droughts and floods occur in different locations.

[0007] Therefore, constructing a novel short-cycle drought-flood abrupt change index based on two factors—soil moisture and precipitation—that balances the rate and degree of abrupt change and is suitable for regional-scale identification is of great scientific significance for improving event identification capabilities, deepening the study of its mechanisms, and improving the accuracy of disaster prediction. Summary of the Invention

[0008] To address the limitations of existing technologies in calculating the drought-flood transition index, such as biased identification of drought processes due to the simplistic use of factors, inconsistent weighting of transition rate and severity, and overly broad or subjective delineation of event areas, this invention proposes a regional drought-flood transition disaster monitoring method that balances transition rate and severity. Based on soil moisture and precipitation, which effectively characterize drought and flood states, a short-cycle drought-flood transition index suitable for regional-scale identification is constructed by balancing the transition rate and severity, significantly improving the accuracy and reliability of drought-flood transition event monitoring.

[0009] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows:

[0010] A method for monitoring regional drought-flood transition disasters where the rate and severity of rapid transition are balanced, the method comprising the following steps:

[0011] S1: Obtain daily soil moisture data and precipitation data of the study area during the study period. Perform weighted processing according to the depth weight of each soil layer in the soil moisture data, and calculate the standardized soil moisture index SSI and standardized precipitation index SPI respectively.

[0012] S2, the standardized soil moisture index (SSI) and standardized precipitation index (SPI) of the study area were processed by moving average to obtain the SSIM and SPIM sequences of the study area, and the drought and flood attributes of each grid point in the corresponding moving window were identified.

[0013] S3. Select the regional average values ​​of the SSIM and SPIM sequences in the study area and perform lead-lag regression to obtain the number of days with the largest correlation coefficient, b. Combine the number of days with the lead to analyze the drought and flood attribute change patterns of each grid point in the study area and screen out the drought and flood transition grid points in the study area.

[0014] S4. During the study period, the identified drought-flood transition points are statistically analyzed to screen out regional drought-flood transition events and their corresponding consecutive dates within the study area.

[0015] S5. Taking the date with the most drought-flood transition points within each consecutive date range as the base date, the mean SPIM and mean SSIM values ​​for each drought-flood transition point within M days before and after the base date are taken, forming a mean sequence for each. The slopes of the two mean sequences, Diff(SPIM) and Diff(SSIM), are obtained by forward differencing, and the ratio of their maximum values ​​is selected as the multiplier parameter. :

[0016] ;

[0017] S6. Taking the first date before the base date where Diff(SPIM) is greater than zero as the occurrence date of the drought-flood transition event, the SPIM and SSIM values ​​of each drought-flood transition event are synthesized N days before and after the occurrence date to obtain the SPIM sequence and SSIM sequence for each drought-flood transition event, combined with the multiplier parameter. The rapid turning rate was calculated. :

[0018] ;

[0019] In the formula, j = 1, 2, ..., b, where b is the number of days leading the SSIM and SPIM correlation coefficients. and These represent the SPIM values ​​for day i and day i+1, respectively. and Let i and j represent the SSIM values ​​for day i+j and day i+j+1, respectively. The index i is the index of the day the drought-flood transition event occurred, and the index j is the index of the number of days preceding the event. , , and All are timestamps; Within the allowed range of values ​​for variable j, calculate the expression. The maximum value;

[0020] For the SSIM and SPIM sequences of each event, the difference between their respective maximum and minimum values ​​is taken as the degree of abrupt change. The calculation formula is as follows:

[0021] ;

[0022] S7. After calculating the rate and degree of abrupt change for all events, normalize them separately and sum them to obtain the short-cycle drought-flood abrupt change index. According to the short-cycle drought-flood transition index Select rapid drought-flood transition events from drought-flood transition events.

[0023] Furthermore, in step S1, the calculation process of the standardized precipitation index (SPI) includes the following steps:

[0024] S11, calculate the 5-day sliding cumulative precipitation from the daily precipitation data for the study period. ;

[0025] S12, the parameter estimation is performed using the maximum likelihood estimation method, and the calculation formula is as follows:

[0026] ;

[0027] ;

[0028] in, Let n be the mean of the cumulative precipitation series, and n be the sample size. For shape parameters, A is the scale parameter, and A is the first intermediate variable used to calculate the shape parameter;

[0029] S13, Calculate the sliding cumulative precipitation probability density function The calculation formula is as follows:

[0030] ;

[0031] in, It is the Gamma function;

[0032] S14, for a given sliding cumulative precipitation The cumulative probability of its Gamma distribution is:

[0033] ;

[0034] In the formula, t represents the integration variable. Let x be the cumulative distribution function of the sliding cumulative precipitation; total cumulative probability. for:

[0035] ;

[0036] The proportion of samples with zero precipitation in the total sample was... , This represents the number of samples where precipitation is zero.

[0037] Total cumulative probability The Z-value, converted to a standard normal distribution, is calculated using the following formula:

[0038] make , when hour:

[0039] ;

[0040] ;

[0041] when hour:

[0042] ;

[0043] ;

[0044] The constant term takes the following values: , , , , , p represents the total cumulative probability. The second intermediate variable when converting to the Z-value of a standard normal distribution.

[0045] Step S2 further includes:

[0046] The standardized soil moisture index (SSI) and standardized precipitation index (SPI) of the study area were subjected to N1-day and N2-day moving averages, respectively, to obtain the SSIM and SPIM sequences of the study area.

[0047] Grid points with SSIM values ​​less than or equal to a preset humidity threshold for N1 consecutive days are identified as drought, and grid points with SPIM values ​​greater than or equal to a preset rainfall threshold for N2 consecutive days are identified as floods.

[0048] Step S3 further includes:

[0049] The regional average values ​​of the SSIM and SPIM sequences in the study area were selected for lead-lag regression to obtain the number of days b with the highest correlation coefficient.

[0050] If a grid point's attribute changes from drought to flood within M days, and the SSIM value on day b+1 after the change to flood attribute is greater than a preset humidity threshold, then the grid point is identified as a drought-to-flood transition grid point.

[0051] Further, in step S4, within the study period and study area, the total number of grid points that will change from drought to flood in the next few days is calculated for each day to obtain a grid point total sequence. The 95th percentile of this grid point total sequence is used as the regional event threshold to filter out all dates that exceed the regional event threshold. Among them, dates that meet the condition of exceeding the regional event threshold for two consecutive days or more are merged into one event to obtain the regional drought-flood transition events and corresponding consecutive dates within the study area.

[0052] Furthermore, in step S7, based on the short-cycle drought-flood transition index... The drought-flood transition events in the study area are ranked, and the drought-flood transition events that are ranked before the preset ranking threshold are identified as drought-flood abrupt transition events.

[0053] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0054] First, the regional drought-flood transition disaster monitoring method of the present invention, which balances the rate and degree of rapid transition, introduces a dual-factor synergistic mechanism of soil moisture and precipitation, and first identifies the drought-to-flood process of a single point at the grid scale. Then, through regional threshold determination and quantitative screening of rapid transition rate and degree, a complete technical route for identifying drought-flood transition events is constructed, thereby significantly improving the accuracy and reliability of drought-flood transition event monitoring.

[0055] Secondly, the regional drought-flood transition disaster monitoring method of the present invention, which balances the rate and severity of rapid transition, uses a short-cycle drought-flood transition index composed of both the rate and severity of rapid transition, with their weights balanced and unified. This allows for a more accurate and effective assessment of the risk level of drought-flood transition events. Unlike traditional methods that use regional averages to calculate the index, the regional threshold and the final screening threshold for drought-flood transition events in this invention can be flexibly adjusted according to the actual climate and geographical characteristics of different regions. Therefore, it has stronger regional adaptability and ensures the comprehensiveness and objectivity of event screening.

[0056] Third, the regional drought-flood transition disaster monitoring method of the present invention, which balances the rate and degree of rapid transition, identifies drought-flood transition events that exhibit significant regional characteristics, providing new insights into the mechanism of drought-flood transition. At the same time, this framework can also be extended to the identification of drought-flood transition events, thereby providing reliable technical support for regional climate change monitoring, disaster formation mechanism analysis, and comprehensive risk assessment. Attached Figure Description

[0057] Figure 1 This is a flowchart of the regional drought-flood transition disaster monitoring method that balances the rate and degree of rapid change according to the present invention.

[0058] Figure 2 A schematic diagram of the leading and lagging regression coefficients of SSIM and SPIM in the Yangtze River Basin;

[0059] Figure 3 A schematic diagram showing the frequency distribution of drought-to-flood transitions in the Yangtze River Basin from 1961 to 2020;

[0060] Figure 4 Composite curves of SPIM and SSIM slopes for 167 drought-to-flood events;

[0061] Figure 5 A schematic diagram showing the temporal distribution of drought-to-flood events in the Yangtze River Basin;

[0062] Figure 6 This is a schematic diagram showing the temporal distribution of rapid shifts between drought and flood in the Yangtze River Basin. Detailed Implementation

[0063] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0064] This invention discloses a method for monitoring regional drought-flood transition disasters that balances the rate and degree of rapid transition. The method includes the following steps:

[0065] S1: Obtain daily soil moisture data and precipitation data of the study area during the study period. Perform weighted processing according to the depth weight of each soil layer in the soil moisture data, and calculate the standardized soil moisture index SSI and standardized precipitation index SPI respectively.

[0066] S2, the standardized soil moisture index (SSI) and standardized precipitation index (SPI) of the study area were processed by moving average to obtain the SSIM and SPIM sequences of the study area, and the drought and flood attributes of each grid point in the corresponding moving window were identified.

[0067] S3. Select the regional average values ​​of the SSIM and SPIM sequences in the study area and perform lead-lag regression to obtain the number of days with the largest correlation coefficient, b. Combine the number of days with the lead to analyze the drought and flood attribute change patterns of each grid point in the study area and screen out the drought and flood transition grid points in the study area.

[0068] S4. During the study period, the identified drought-flood transition points are statistically analyzed to screen out regional drought-flood transition events and their corresponding consecutive dates within the study area.

[0069] S5. Taking the date with the most drought-flood transition points within each consecutive date range as the base date, the mean SPIM and mean SSIM values ​​for each drought-flood transition point within M days before and after the base date are taken, forming a mean sequence for each. The slopes of the two mean sequences, Diff(SPIM) and Diff(SSIM), are obtained by forward differencing, and the ratio of their maximum values ​​is selected as the multiplier parameter. :

[0070] ;

[0071] S6. Taking the first date before the base date where Diff(SPIM) is greater than zero as the occurrence date of the drought-flood transition event, the SPIM and SSIM values ​​of each drought-flood transition event are synthesized N days before and after the occurrence date to obtain the SPIM sequence and SSIM sequence for each drought-flood transition event, combined with the multiplier parameter. The rapid turning rate was calculated. :

[0072] ;

[0073] In the formula, j = 1, 2, ..., b, where b is the number of days leading the SSIM and SPIM correlation coefficients. and These represent the SPIM values ​​for day i and day i+1, respectively. and Let i and j represent the SSIM values ​​for day i+j and day i+j+1, respectively. The index i is the index of the day the drought-flood transition event occurred, and the index j is the index of the number of days preceding the event. , , and All are timestamps; Within the allowed range of values ​​for variable j, calculate the expression. The maximum value;

[0074] For the SSIM and SPIM sequences of each event, the difference between their respective maximum and minimum values ​​is taken as the degree of abrupt change. The calculation formula is as follows:

[0075] ;

[0076] S7. After calculating the rate and degree of abrupt change for all events, normalize them separately and sum them to obtain the short-cycle drought-flood abrupt change index. According to the short-cycle drought-flood transition index Select rapid drought-flood transition events from drought-flood transition events.

[0077] See Figure 1 The monitoring method of the present invention specifically includes the following steps (the specific number of days in the following embodiments is only for the purpose of illustrating the principle of the scheme. In practical applications, it can be set according to the geographical environment and climate characteristics of the study area):

[0078] 1) Data Acquisition

[0079] Acquire daily soil moisture and precipitation data for a specified range within a given time period, and perform weighted processing based on the depth weights of each soil layer in the soil moisture data. Both datasets must have the same resolution.

[0080] 2) Calculate the standardized soil moisture index and the standardized precipitation index.

[0081] Based on the daily soil moisture and precipitation data of the study area obtained in step 1), the standardized soil moisture index (SSI) and standardized precipitation index (SPI) are calculated respectively. Taking the calculation of SPI as an example, the calculation method of SSI is the same as that of SPI.

[0082] 2.1) Calculate the 5-day moving cumulative precipitation for the daily precipitation data during the study period. .

[0083] 2.2) Perform Gamma distribution fitting. First, estimate the parameters using the maximum likelihood estimation method. The calculation formula is as follows:

[0084] (1)

[0085] (2)

[0086] in, Let n be the mean of the cumulative precipitation series, and n be the sample size. For shape parameters, Let A be the scale parameter, and let A be the first intermediate variable used to calculate the shape parameter.

[0087] 2.3) Calculate the sliding cumulative precipitation probability density function The calculation formula is as follows:

[0088] (3)

[0089] in, This is the Gamma function.

[0090] 2.4) Calculate the cumulative probability. For a given sliding cumulative precipitation... The cumulative probability of its Gamma distribution is:

[0091] (4)

[0092] Considering that precipitation may be zero, the total cumulative probability for:

[0093] (5)

[0094] The proportion of samples with zero precipitation in the total sample was... , This represents the number of samples where precipitation is zero.

[0095] 2.5) Standardization Conversion

[0096] Total cumulative probability The Z-value, or SPI value, is converted to a standard normal distribution. The probability integral transformation here uses the standard method recommended by the World Meteorological Organization, as shown in the following formula:

[0097] make ,when hour:

[0098] (6)

[0099] (7)

[0100] when hour:

[0101] (8)

[0102] (9)

[0103] The constant term takes the following values: , , , , , p represents the total cumulative probability. The second intermediate variable when converting to the Z-value of a standard normal distribution.

[0104] 3) Screening of single-point drought-to-flood transition processes

[0105] To effectively suppress high-frequency fluctuations in the daily sequences and highlight abrupt shifts between drought and flood, 5-day and 15-day moving averages were applied to the SPIM and SSIM sequences, respectively, to obtain smoothed SPIM and SSIM sequences. At each grid point, SSIM values ​​less than or equal to -0.5 for 15 consecutive days were identified as drought, and SPIM values ​​greater than or equal to 0.5 for 5 consecutive days were identified as flood.

[0106] The regional averages of SSIM and SPIM for the study area were selected, and their lead-lag regression coefficients were calculated. The number of days leading with the highest correlation coefficient was [value missing]. If the grid points achieve a transition from drought to flood within 5 days, and the flood begins on the [number]th day... If SSIM > -0.5, it is identified as a single-point drought-to-flood process.

[0107] 4) Screening for regional drought-flood transition events

[0108] Within the study period and region, the total number of grid points that would transition from drought to flood within the next 5 days was calculated for each day. The 95th percentile of this grid point count sequence was used as the regional event threshold, and all dates exceeding this threshold were selected. Dates that met the regional event threshold condition for two or more consecutive days were combined into a single event.

[0109] 5) Determine the turning point date

[0110] Within each consecutive date range, using the date with the most drought-to-flood grid points as the base date, calculate the mean values ​​of SPIM and SSIM for the 20 days before and after the occurrence of the drought-to-flood grid point within the next 5 days for all events. This yields two sequences, each of length 41. The forward difference between the two sequences represents the slope of the changes in SPIM and SSIM. and Select The first date with a value greater than zero before the base date is taken as the date of occurrence of the drought-to-flood event, i.e., day 0.

[0111] 6) Construct a new short-cycle drought-flood transition index

[0112] New short-cycle drought-flood rapid transition index It consists of the abrupt change rate and the abrupt change severity. First, the SPIM and SSIM values ​​for each event are synthesized 20 days before and after the drought-to-flood grid point, resulting in the SPIM and SSIM sequences for each event. These sequences are used to calculate the abrupt change rate and severity for each event. Based on the synthesized values ​​of all events calculated in step 5),... and For sequences, the ratio of the maximum values ​​of the two sequences is taken as the multiplier parameter. Rapid turning rate and multiple parameters The calculation formula is as follows:

[0113] (10)

[0114] (11)

[0115] In the formula, j = 1, 2, ..., b; then, the degree of abrupt change is determined. For each SSIM and SPIM sequence of length 41, the difference between their maximum and minimum values ​​is taken as the degree of abrupt change. The calculation formula is as follows:

[0116] (12)

[0117] 7) Screening for sudden shifts between drought and flood events

[0118] After calculating the rate and degree of abrupt change for all events, the two are normalized separately and then summed to obtain the novel short-cycle drought-flood abrupt change index. Events ranking in the top 50% of the drought-flood abrupt change index are considered drought-flood abrupt change events.

[0119] The following example, using the defined DFAAI to screen short-cycle drought-flood transition events in the Yangtze River basin, further illustrates the specific implementation steps and parameter settings of this invention with reference to the accompanying drawings, ensuring that those skilled in the art can fully reproduce it. See the technical flowchart below. Figure 1 .

[0120] 1) The Yangtze River Basin (24°–36°N, 90°–123°E) was selected as the study area. Hourly soil moisture data from 1961 to 2020 were downloaded from the European Centre for Medium-Range Weather Forecasts (ECMWF) 5th Generation Atmospheric Reanalysis dataset. The depth of the first three layers (0-100cm) was selected. , , The spatial resolution is 0.25°×0.25°. Precipitation data is from the China Meteorological Administration's CN05.1 gridded daily dataset, also with a spatial resolution of 0.25°×0.25°. Soil moisture data is processed into daily data and weighted according to the depth of each soil layer. .

[0121] 2) Extract daily soil and precipitation data for the Yangtze River basin. Using the method described in step 2), calculate the SPI from precipitation data and the SSI from soil data. First, calculate the 5-day moving cumulative precipitation for each grid point from 1961 to 2020. The parameters are calculated based on formulas (1) and (2) in step 2 of the technical solution. and Find the values ​​in formulas (3) and (4). The probability density distribution function and the cumulative probability of the Gamma distribution Then, according to formula (5), Find the total cumulative probability Finally, according to formula (6) Formula (10) is standardized to obtain SPI and SSI.

[0122] 3) To suppress high-frequency disturbances, the SPI and SSI sequences were processed with 5-day and 15-day moving averages, respectively, to obtain smoothed SPIM and SSIM sequences. Drought was defined as SSMI less than or equal to -0.5 for 15 consecutive days, and flooding as SPIM greater than or equal to 0.5 for 5 consecutive days. Lead-lag regression was performed on the regional averages of SSIM and SPIM in the Yangtze River Basin, revealing that SPIM changes lead SSIM changes by 7 days. Figure 2 As shown, that is Therefore, if a grid point achieves a drought-to-flood transition within 5 days, and the SSIM value is greater than -0.5 on the 8th day after the start of the flood, it is identified as a single-point drought-to-flood transition process. Figure 2 The horizontal axis represents the number of days leading / lagging, with negative values ​​indicating leading days and positive values ​​indicating lagging days. The vertical axis represents the linear correlation between the standardized soil moisture mean series (SSIM) and the standardized precipitation mean series (SPIM) under different leading / lagging days. The horizontal dashed line is the baseline within the current statistical day range; above the baseline indicates a positive correlation, and below it indicates a negative correlation. Based on the above determinations, the frequency of drought-to-flood transitions at each grid point within the study area from 1961 to 2020 can be statistically analyzed, and its spatial distribution is shown below. Figure 3 As shown.

[0123] 4) Calculate the total number of grid points in the Yangtze River Basin from 1961 to 2020 that indicate a drought-to-flood transition within the next 5 days. The 95th percentile of this grid point sequence is 163 grid points, which is used as the threshold for regional events in the Yangtze River Basin. All dates exceeding the threshold are selected, and dates with two or more consecutive days are combined into one event, resulting in a total of 167 drought-to-flood transition events.

[0124] 5) Next, within each consecutive date segment, the date with the most drought-to-flood grid points is selected as the baseline date. Within the next 5 days of each baseline date from the 167 events, the SPIM and SSIM sequences for the 20 days before and after the drought-to-flood grid point are synthesized, resulting in SPIM and SSIM sequences of length 41 days each. The forward difference between the two sequences represents the slope of the SPIM and SSIM changes. and ,like Figure 4 As shown, Figure 4 The horizontal axis represents the number of days leading / lagging, and the vertical axis represents the slope of change. The base date is selected. The first date greater than zero is considered the turning point from drought to flood, i.e., day 0. The date distribution of drought-to-flood events in the Yangtze River basin is as follows: Figure 5 As shown, the bar chart represents the number of drought-to-flood events that occurred in the corresponding month, and the curve is the interdecadal trend line of the number of drought-to-flood events in the Yangtze River Basin in each month.

[0125] 6) Next, calculate the abrupt change rate and degree of abrupt change for each event. Select the 5-day period following the turning point of each event, and synthesize the SPIM and SSIM sequences for the 20 days before and after the drought-to-flood grid point, resulting in 167 SPIM and SSIM sequences. Calculate the multiplier parameter according to formula (10) in step 6) of the technical solution. .Will , Substituting the SPIM and SSIM sequences of each event into formulas (11) and (12) yields the abrupt change rate and abrupt change degree for each event. After normalizing the abrupt change rate and abrupt change degree of all events and summing them, a novel short-period drought-flood abrupt change index for each event is obtained. .

[0126] 7) Finally, by selecting the events ranking in the top 50% of the drought-flood abrupt transition index, we can obtain 83 drought-flood abrupt transition events in the Yangtze River Basin from 1961 to 2020. The date distribution of these events is as follows: Figure 6 As shown, the bars represent the number of drought-to-flood abrupt transition events occurring in the corresponding month, and the curve represents the characteristic value of the drought-to-flood abrupt transition index (DFAAI) for the corresponding month, reflecting the intensity level of the drought-to-flood abrupt transition events. The fluctuations of the curve correspond to the intensity changes of drought-to-flood abrupt transition events in each month. Compared to drought-to-flood events, the screened drought-to-flood abrupt transition events are mainly concentrated in May, June, and August, which is consistent with the actual situation in the Yangtze River Basin. Furthermore, the events screened are more comprehensive than those in existing studies, further demonstrating the objectivity and reliability of the method.

[0127] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0128] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for monitoring regional drought-flood transition disasters with a balance between the rate and degree of rapid transition, characterized in that, The method includes the following steps: S1: Obtain daily soil moisture data and precipitation data of the study area during the study period. Perform weighted processing according to the depth weight of each soil layer in the soil moisture data, and calculate the standardized soil moisture index SSI and standardized precipitation index SPI respectively. S2, the standardized soil moisture index (SSI) and standardized precipitation index (SPI) of the study area were processed by moving averages of N1 days and N2 days, respectively, to obtain the SSIM and SPIM sequences of the study area. Grid points with SSIM values ​​less than or equal to a preset humidity threshold for N1 consecutive days are identified as drought, and grid points with SPIM values ​​greater than or equal to a preset rainfall threshold for N2 consecutive days are identified as flood. S3, select the regional average values ​​of the SSIM sequence and SPIM sequence of the study area to perform lead-lag regression, and obtain the number of days b with the largest correlation coefficient; if the grid point attribute changes from drought to flood within M days, and the SSIM value on the (b+1)th day after the change to flood attribute is greater than the preset humidity threshold, then the grid point is identified as a drought-flood transition grid point. S4. During the study period and within the study area, calculate the total number of grid points that will turn from drought to flood in the next few days for each day to obtain a grid point total sequence. Use the 95th percentile of this grid point total sequence as the regional event threshold to filter out all dates that exceed the regional event threshold. Among them, dates that meet the condition of exceeding the regional event threshold for two consecutive days or more are merged into one event to filter out the regional drought-flood transition events and corresponding consecutive dates within the study area. S5. Taking the date with the most drought-flood transition points within each consecutive date range as the base date, the mean SPIM and mean SSIM values ​​for each drought-flood transition point within M days before and after the base date are taken, forming a mean sequence for each. The slopes of the two mean sequences, Diff(SPIM) and Diff(SSIM), are obtained by forward differencing, and the ratio of their maximum values ​​is selected as the multiplier parameter. : ； S6. Taking the first date before the base date where Diff(SPIM) is greater than zero as the occurrence date of the drought-flood transition event, the SPIM and SSIM values ​​of each drought-flood transition event are synthesized N days before and after the occurrence date to obtain the SPIM sequence and SSIM sequence for each drought-flood transition event, combined with the multiplier parameter. The rapid turning rate was calculated. : ； In the formula, j = 1, 2, ..., b, where b is the number of days leading the SSIM and SPIM correlation coefficients. and These represent the SPIM values ​​for day i and day i+1, respectively. and Let i and j represent the SSIM values ​​for day i+j and day i+j+1, respectively. The index i is the index of the day the drought-flood transition event occurred, and the index j is the index of the number of days preceding the event. , , and All are timestamps; Within the allowed range of values ​​for variable j, calculate the expression. The maximum value; For the SSIM and SPIM sequences of each event, the difference between their respective maximum and minimum values ​​is taken as the degree of abrupt change. The calculation formula is as follows: ； S7. After calculating the rate and degree of abrupt change for all events, normalize them separately and sum them to obtain the short-cycle drought-flood abrupt change index. According to the short-cycle drought-flood transition index Select rapid drought-flood transition events from drought-flood transition events.

2. The regional drought-flood transition disaster monitoring method with balanced transition rate and degree as described in claim 1, characterized in that, In step S1, the calculation process of the Standardized Precipitation Index (SPI) includes the following steps: S11, calculate the 5-day sliding cumulative precipitation from the daily precipitation data for the study period. ; S12, the parameter estimation is performed using the maximum likelihood estimation method, and the calculation formula is as follows: ； ； in, Let n be the mean of the cumulative precipitation series, and n be the sample size. For shape parameters, A is the scale parameter, and A is the first intermediate variable used to calculate the shape parameter; S13, Calculate the sliding cumulative precipitation probability density function The calculation formula is as follows: ； in, It is the Gamma function; S14, for a given sliding cumulative precipitation The cumulative probability of its Gamma distribution is: ； In the formula, t represents the integration variable. Let x be the cumulative distribution function of the sliding cumulative precipitation; total cumulative probability. for: ； The proportion of samples with zero precipitation in the total sample was... , This represents the number of samples where precipitation is zero. Total cumulative probability The Z-value, converted to a standard normal distribution, is calculated using the following formula: make , when hour: ； ； when hour: ； ； The constant term takes the following values: , , , , , p represents the total cumulative probability. The second intermediate variable when converting to the Z-value of a standard normal distribution.

3. The regional drought-flood transition disaster monitoring method with balanced transition rate and degree as described in claim 1, characterized in that, In step S7, based on the short-cycle drought-flood transition index... The drought-flood transition events in the study area are ranked, and the drought-flood transition events that are ranked before the preset ranking threshold are identified as abrupt drought-flood transition events.

Images