Method for identifying multi-scale drought-flood abrupt transition probability based on ensemble empirical mode decomposition

Abstract

The application discloses a multi-scale drought-flood abrupt change probability identification method based on ensemble empirical mode decomposition, and constructs a general identification framework suitable for different scale drought-flood abrupt change events of meteorology, agriculture, hydrology and the like. The method comprises the following steps: calculating a daily sliding standardized drought index, and performing ensemble empirical mode decomposition on the standardized drought index to obtain a plurality of intrinsic mode functions representing different time scale signals and a residual component; constructing a frequency feature spectrum of each intrinsic mode function, and synthesizing and calculating the standardized intrinsic mode functions containing week-to-year scale signals based on the frequency feature spectrum; performing drought identification on the synthesized, standardized and decomposed drought index according to drought conditions; performing flood identification on hydro-meteorological variables according to flood conditions; pairing the identified drought and flood events, calculating the time of each drought-flood abrupt change, fitting the drought-flood abrupt change time sequence, and calculating the occurrence probability of different drought-flood abrupt change times through a probability distribution function.

Description

Technical Field

[0001] This invention belongs to the field of natural disaster monitoring and risk assessment technology, particularly the field of monitoring technology for extreme complex events, and specifically relates to a multi-scale drought-flood transition probability identification method based on ensemble empirical mode decomposition. Background Technology

[0002] Climate change and intensive human activities have not only increased the frequency and intensity of individual extreme hydrological and meteorological disasters such as global heat waves, floods, and droughts, but have also significantly increased the frequency and severity of compound extreme events. Rapid shifts between drought and flood, as a typical compound extreme event, have an impact far greater than droughts or floods occurring alone.

[0003] Drought-flood abrupt transition events are generally defined as the rapid shift between drought and flood states within a short period. Therefore, the determination of drought-flood abrupt transitions must be able to identify both the occurrence and development characteristics of drought and flood, as well as the turning point from drought to flood. The identification factors for drought-flood abrupt transitions are mainly divided into precipitation-based indicators and comprehensive hydro-meteorological indicators. Their identification methods can also be categorized into two main types: drought-flood abrupt transition indices based on standardized meteorological and hydrological elements, and threshold methods based on runs theory. Drought-flood abrupt transition indices reflect the intensity of the transition by calculating the difference in hydro-meteorological elements at adjacent time steps, such as the long / short cycle drought-flood abrupt transition index (LDFAI / SDFAI) and the daily drought-flood abrupt transition index (DWAAI). These methods typically use monthly or seasonal timescales and rely on a single threshold to determine drought and flood, failing to accurately capture the turning point or quantify the speed of drought-flood alternation. Furthermore, the empirical coefficients in the drought-flood abrupt transition index calculation formula lack regional universality. Thresholding methods based on runs theory focus more on the entire process of abrupt drought-flood transitions. They commonly use diurnal hydrometeorological elements such as the Standardized Precipitation Index (SPI), Standardized Precipitation Evapotranspiration Index (SPEI), Standardized Runoff Index (SRI), and Weighted Average Precipitation Index (SWAP). By setting drought thresholds, flood thresholds, drought duration, and drought-flood transition time, they emphasize the "rapid" nature of abrupt drought-flood transitions. However, there is currently no unified standard for setting these thresholds. In practical applications, it has been found that for some cases where drought is briefly relieved by a small amount of precipitation and then quickly returns to drought, the drought identified by diurnal indices such as SPI, SPEI, and SRI will be discontinuous, resulting in some drought processes being missed.

[0004] Therefore, how to solve the problems of missed drought detection and large spatial heterogeneity of threshold values ​​is an urgent issue that needs to be considered in order to achieve standardized monitoring and early warning of rapid shifts between drought and flood. Summary of the Invention

[0005] Purpose of the invention: To address the problems and shortcomings of existing technologies, this invention provides a multi-scale drought-flood transition probability identification method based on ensemble empirical mode decomposition. This method solves the problem of temporal discontinuity when traditional monthly or even annual drought indices are directly used for daily-scale drought identification. It also solves the problems of inconsistent judgment criteria and large spatial heterogeneity in existing drought-flood transition identification methods. A general identification framework applicable to multi-scale drought-flood transition events is constructed, providing practical technical support for improving drought-flood transition monitoring, early warning, risk identification, and assessment.

[0006] Technical solution: A multi-scale drought-flood transition probability identification method based on ensemble empirical mode decomposition, comprising the following steps: S1. Calculate the daily sliding standardized drought index and perform ensemble empirical mode decomposition on the standardized drought index to obtain multiple intrinsic mode functions and a residual component characterizing signals at different time scales. S2. Construct the frequency feature map of each intrinsic mode function, select intrinsic mode functions containing periodic-year scale signals based on the frequency feature map, synthesize them, and standardize the synthesis results according to the following formula: (1) Where S is the synthesis of intrinsic mode functions, The average value of the S-sequence is... denoted as the standard deviation of the S sequence.

[0007] S3. Based on drought conditions, identify drought indices that have been decomposed, synthesized, and standardized; S4. Identify floods based on hydrological and meteorological variables according to flood conditions; S5. Pair the identified drought and flood events, calculate the time of drought-to-flood transition for each pair, fit the drought-to-flood time series, and calculate the probability of occurrence of different drought-to-flood transition times using the probability distribution function.

[0008] Furthermore, the standardized drought index in step S1 includes the standardized precipitation index, the standardized soil moisture index, or the standardized runoff index. The formula for calculating the standardized drought index SDI is: (2) (3) (4) in This represents the cumulative hydrological and meteorological variables over m days. This represents the hydrometeorological variables for day ti, including precipitation, soil moisture, or runoff. express cumulative distribution function , Indicates the standardized drought index, The inverse cumulative distribution function represents the standard normal distribution.

[0009] Furthermore, the frequency characteristic spectrum in step S2 is constructed using the Hilbert phase-frequency method, which includes first performing a Hilbert transform on each intrinsic mode function (IMF) to obtain an analytical signal; then extracting the instantaneous amplitude and instantaneous phase; expanding the phase and calculating the instantaneous frequency; filtering instantaneous frequencies with amplitudes greater than 30% by percentiles; calculating the average frequency; and calculating the average period.

[0010] Furthermore, the drought conditions in step S3 are a drought index of less than -1 and a duration of more than 10 days.

[0011] Furthermore, the flood conditions in step S4 are: 3-day cumulative precipitation (or runoff), or 3-day average soil moisture greater than 99.5% of the cumulative frequency distribution value; 5-day cumulative precipitation (or runoff), or 5-day average soil moisture greater than 99.3% of the cumulative frequency distribution value; or 10-day cumulative precipitation (or runoff), or 10-day average soil moisture greater than 98.7% of the cumulative frequency distribution value.

[0012] A multi-scale drought-flood transition probability identification system based on ensemble empirical mode decomposition includes the following modules: Decomposition module: used to calculate the daily sliding standardized drought index and perform ensemble empirical mode decomposition on the standardized drought index to obtain multiple intrinsic mode functions and a residual component characterizing the signal at different time scales. Synthesis module: Used to construct the frequency feature map of each intrinsic mode function, select intrinsic mode functions containing periodic-year scale signals based on the frequency feature map, synthesize them, and standardize the synthesis results according to the following formula: (1) Where S is the synthesis of intrinsic mode functions, The average value of the S-sequence is... The standard deviation of the S sequence; Drought identification module: used to identify drought based on the decomposed, synthesized, and standardized drought index; Flood identification module: Identifies floods based on hydrological and meteorological variables according to flood conditions; Probability calculation module: used to pair identified drought and flood events, calculate the time of drought-to-flood transition for each pair, fit the drought-to-flood time series, and calculate the probability of occurrence of different drought-to-flood transition times through probability distribution functions.

[0013] The implementation process and methods of the system are the same and will not be described again.

[0014] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the multi-scale drought-flood transition probability identification method based on set empirical mode decomposition as described above.

[0015] A computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implements the steps of the multi-scale drought-flood transition probability identification method based on set empirical mode decomposition as described above.

[0016] Beneficial effects: Compared with existing technologies, the multi-scale drought-flood transition probability identification method based on ensemble empirical mode decomposition provided by this invention has the following advantages: (1) This invention solves the problem of time discontinuity when traditional monthly or even annual drought indices are directly used for daily drought identification by combining empirical mode decomposition and synthesis.

[0017] (2) This invention solves the problems of inconsistent judgment criteria and large spatial heterogeneity in existing drought-flood transition identification methods by using a probability-based identification method.

[0018] (3) The present invention constructs a universal identification system applicable to different types of drought and flood sudden change events such as meteorology, agriculture, and hydrology, providing support for drought and flood sudden change risk prevention and management decision-making. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the steps of a multi-scale drought-flood transition probability identification method based on set empirical mode decomposition according to an embodiment of the present invention. Figure 2 This is the upper half of the ensemble empirical mode decomposition of the standardized drought index in the embodiments of the present invention; Figure 3 This is the lower half of the ensemble empirical mode decomposition diagram of the standardized drought index in the embodiments of the present invention; Figure 4 This is a frequency characteristic spectrum of the intrinsic mode functions in an embodiment of the present invention; Figure 5 This is a probability map of the sudden shift between drought and flood identified in the embodiments of the present invention. Detailed Implementation

[0020] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. After reading the present invention, any modifications of the present invention in various equivalent forms by those skilled in the art will fall within the scope defined by the appended claims.

[0021] This invention uses the identification of abrupt drought-flood transitions in the central urban area of ​​Beijing from 1961 to 2022 as an example to provide a multi-scale probability identification method for abrupt drought-flood transitions based on ensemble empirical mode decomposition. Figure 1 As shown, it includes the following steps: S1. Calculate the daily sliding standardized drought index and perform ensemble empirical mode decomposition on it to obtain multiple intrinsic mode functions and a residual component characterizing the signal at different time scales, such as... Figure 2-3 As shown, from top to bottom, the sequence is: original sequence, intrinsic mode function 1, intrinsic mode function 2, ..., intrinsic mode function 14.

[0022] In this embodiment of the invention, the Standardized Precipitation Index (SPI) is selected as the drought index, and the formula for daily sliding calculation of the SPI is: (1) (2) (3) in This represents the cumulative hydrometeorological variable (precipitation, soil moisture, or runoff) over m days. In this implementation case, m is 30 days, and the hydrometeorological variable is precipitation. express The cumulative distribution function is fitted using the gamma distribution in this implementation example. This represents the standardized drought index, which is SPI in this implementation case.

[0023] S2. Construct the frequency feature map of each intrinsic mode function, select intrinsic mode functions containing periodic and annual scale signals based on the frequency feature map, synthesize them, and standardize them. The frequency characteristic spectrum is constructed using the Hilbert phase-frequency method, which includes first performing a Hilbert transform on each intrinsic mode function (IMF) to obtain the analytic signal; then extracting the instantaneous amplitude and instantaneous phase; expanding the phase and calculating the instantaneous frequency; filtering instantaneous frequencies with amplitudes greater than 30% using quantiles; calculating the average frequency; and calculating the average period, such as... Figure 4 As shown. In this embodiment, the cyclical annual scale signal is represented by the intrinsic mode functions 3-10, which are summed and standardized.

[0024] S3. Based on drought conditions, identify drought indices that have been decomposed, synthesized, and standardized; The drought condition is defined as a drought index less than -1 that lasts for more than 10 days.

[0025] S4. Identify floods based on hydrological and meteorological variables according to flood conditions; The flood conditions are defined as follows: 3-day cumulative precipitation (or runoff), or 3-day average soil moisture greater than 99.5% of the cumulative frequency distribution value; 5-day cumulative precipitation (runoff), or 5-day average soil moisture greater than 99.3% of the cumulative frequency distribution value; or 10-day cumulative precipitation (or runoff), or 10-day average soil moisture greater than 98.7% of the cumulative frequency distribution value. In this embodiment, 3-day cumulative precipitation, 5-day cumulative precipitation, and 10-day cumulative precipitation are selected for flood identification. When making a judgment, if any one of the three conditions is met, it is identified as flooding. However, all three conditions need to be considered when making a judgment. Only one of precipitation, soil moisture, and runoff needs to be selected in a single instance.

[0026] S5. Pair the identified drought and flood events, calculate the time of each drought-to-flood transition, fit the drought-to-flood time series, and calculate the probability of occurrence of different drought-to-flood transition times using the probability distribution function.

[0027] In this embodiment, a generalized extreme value is used to fit the drought-to-flood time series. The probability results of different abrupt transition times are as follows: Figure 5 As shown.

[0028] A multi-scale drought-flood transition probability identification system based on ensemble empirical mode decomposition includes the following modules: Decomposition module: used to calculate the daily sliding standardized drought index and perform ensemble empirical mode decomposition on the standardized drought index to obtain multiple intrinsic mode functions and a residual component characterizing the signal at different time scales. Synthesis module: Used to construct the frequency feature map of each intrinsic mode function, select intrinsic mode functions containing periodic-year scale signals based on the frequency feature map, synthesize them, and standardize the synthesis results according to the following formula: (1) Where S is the synthesis of intrinsic mode functions, The average value of the S-sequence is... The standard deviation of the S sequence; Drought identification module: used to identify drought based on the decomposed, synthesized, and standardized drought index; Flood identification module: Identifies floods based on hydrological and meteorological variables according to flood conditions; Probability calculation module: used to pair identified drought and flood events, calculate the time of drought-to-flood transition for each pair, fit the drought-to-flood time series, and calculate the probability of occurrence of different drought-to-flood transition times through probability distribution functions.

[0029] Obviously, those skilled in the art should understand that the steps of the multi-scale drought-flood transition probability identification method based on set empirical mode decomposition or the modules of the multi-scale drought-flood transition probability identification system based on set empirical mode decomposition in the above embodiments of the present invention can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented here, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the embodiments of the present invention are not limited to any particular hardware and software combination.

[0030] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited thereto. Any simple changes or equivalent substitutions of the technical solutions that can be obviously obtained by those skilled in the art within the scope of the technology disclosed in the present invention shall fall within the scope of protection of the present invention.

Claims

1. A multi-scale drought-flood transition probability identification method based on ensemble empirical mode decomposition, characterized in that, Includes the following steps: S1. Calculate the daily sliding standardized drought index and perform ensemble empirical mode decomposition on the standardized drought index to obtain multiple intrinsic mode functions and a residual component characterizing signals at different time scales. S2. Construct the frequency feature map of each intrinsic mode function, select intrinsic mode functions containing periodic-year scale signals based on the frequency feature map, synthesize them, and standardize the synthesis results according to the following formula: （1） Where S is the synthesis of intrinsic mode functions, The average value of the S-sequence. The standard deviation of the S sequence; S3. Based on drought conditions, identify drought indices that have been decomposed, synthesized, and standardized; S4. Identify floods based on hydrological and meteorological variables according to flood conditions; S5. Pair the identified drought and flood events, calculate the time of drought-to-flood transition for each pair, fit the drought-to-flood time series, and calculate the probability of occurrence of different drought-to-flood transition times using the probability distribution function.

2. The multi-scale drought-flood transition probability identification method based on ensemble empirical mode decomposition according to claim 1, characterized in that, The standardized drought index in step S1 includes the standardized precipitation index, standardized soil moisture index, or standardized runoff index. The formula for calculating the standardized drought index SDI is: (2) (3) (4) in This represents the cumulative hydrological and meteorological variables over m days. This represents the hydrometeorological variables for day ti, including precipitation, soil moisture, or runoff. express cumulative distribution function , Indicates the standardized drought index, The inverse cumulative distribution function represents the standard normal distribution.

3. The multi-scale drought-flood transition probability identification method based on ensemble empirical mode decomposition according to claim 1, characterized in that, The frequency characteristic spectrum in step S2 is constructed using the Hilbert phase-frequency method, which includes first performing a Hilbert transform on each intrinsic mode function (IMF) to obtain an analytical signal; then extracting the instantaneous amplitude and instantaneous phase; expanding the phase and calculating the instantaneous frequency; filtering instantaneous frequencies with amplitudes greater than 30% by percentiles; calculating the average frequency; and calculating the average period.

4. The multi-scale drought-flood transition probability identification method based on ensemble empirical mode decomposition according to claim 1, characterized in that, The drought conditions in step S3 are a drought index less than -1 and lasting for more than 10 days.

5. The multi-scale drought-flood transition probability identification method based on ensemble empirical mode decomposition according to claim 1, characterized in that, The flood conditions in step S4 are: 3-day cumulative precipitation (or runoff), or 3-day average soil moisture greater than 99.5% of the cumulative frequency distribution value; 5-day cumulative precipitation (or runoff), or 5-day average soil moisture greater than 99.3% of the cumulative frequency distribution value; or 10-day cumulative precipitation (or runoff), or 10-day average soil moisture greater than 98.7% of the cumulative frequency distribution value.

6. A multi-scale drought-flood transition probability identification system based on ensemble empirical mode decomposition, characterized in that, Includes the following modules: Decomposition module: used to calculate the daily sliding standardized drought index and perform ensemble empirical mode decomposition on the standardized drought index to obtain multiple intrinsic mode functions and a residual component characterizing the signal at different time scales. Synthesis module: Used to construct the frequency feature map of each intrinsic mode function, select intrinsic mode functions containing periodic-year scale signals based on the frequency feature map, synthesize them, and standardize the synthesis results according to the following formula: （1） Where S is the synthesis of intrinsic mode functions, The average value of the S-sequence is... The standard deviation of the S sequence; Drought identification module: used to identify drought based on the decomposed, synthesized, and standardized drought index; Flood identification module: Identifies floods based on hydrological and meteorological variables according to flood conditions; Probability calculation module: used to pair identified drought and flood events, calculate the time of drought-to-flood transition for each pair, fit the drought-to-flood time series, and calculate the probability of occurrence of different drought-to-flood transition times through probability distribution functions.

7. A computer device, characterized in that: The computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the multi-scale drought-flood transition probability identification method based on set empirical mode decomposition as described in any one of claims 1-5.

8. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that: When the computer program / instruction is executed by the processor, it implements the steps of the multi-scale drought-flood transition probability identification method based on set empirical mode decomposition as described in any one of claims 1-5.

Images