Drainage basin later flood season dynamic identification method

Abstract

The invention provides a watershed later flood season dynamic identification method, which comprises the steps of collecting historical daily average water level, flow and other water regimen data of each hydrological station in a research area, determining the numerical value range of each water regimen factor, and judging whether plums emerge in a watershed or not; generating a plurality of groups of random numbers obeying uniform distribution, and further randomly generating a plurality of groups of water regimen factor thresholds; selecting basin type and regional type typical flood season processes with representativeness in history as a test data set; determining the earliest time when the water level of each main control station in the drainage basin is lower than the warning water level and does not rise later according to the inspection set, and taking the latest time in each station as the flood-control-risk-free time of the drainage basin; and comparing each group of randomly generated water regimen factor threshold values with a typical annual flood season process day by day, and when the water regimen factors of all stations are continuously lower than the corresponding threshold values, judging that the next flood season is entered, thereby providing a new analysis view angle and technical support for flood prevention and drought resistance decision-making of the large river basin.

Description

Technical Field

[0001] This invention relates to the field of flood season grading, and more particularly to a method for dynamic identification of the post-flood season in a river basin. Background Technology

[0002] Due to the seasonal differences in floods, the flood season can be further subdivided into several phases. These phases, along with their corresponding design floods, provide crucial scientific basis for reservoirs to adopt phased flood control level scheduling. Traditional flood season phased methods typically rely on mathematical and statistical analysis of historical flood samples, analyzing and determining the seasonality of flood samples (such as annual maximum flood samples) to derive fixed main flood season and post-flood season phased nodes. These methods are well-suited for watersheds with a single flood source and clear temporal patterns, but they have significant limitations in watersheds with complex flood-prone areas.

[0003] In flood-prone areas, the watersheds are complex, with significant differences in the start and end times of the flood season, the sources of floods, and their combinations among the upstream main streams and tributaries.

[0004] Traditional methods often only allow for the determination of uniform phased points based on a specific type of dominant flood, making it difficult to accurately reflect the asynchronous receding patterns of floods from different sources and their dynamic impact on overall flood control risk. This simplified approach may lead to biases in post-flood season identification—early identification could cause reservoirs to lower their flood control standards prematurely, increasing later flood risks; late identification could unnecessarily prolong the high-water-level operation period, limiting the reservoir's beneficial storage capacity utilization at the end of the flood season. Taking the middle reaches of the Yangtze River as an example, its floods receive water from multiple sources, including the upstream main stream, the Qingjiang River, and the Dongting Lake system. The main flood seasons of these floods differ in timing: the main flood season for the Dongting Lake system is from May to July, while the main flood season for the upper reaches of the Yangtze River is from early June to mid-September.

[0005] Therefore, for watersheds with complex flood compositions, there is an urgent need to develop a dynamic identification method for the post-flood season that can adapt to the diverse sources and staggered timing of floods, so as to more accurately support the flood control scheduling of reservoirs and the efficient utilization of water resources. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of the prior art by providing a dynamic identification method for the post-flood season of a river basin, which can provide dynamic identification for the post-flood season of large river basins with complex flood-prone areas.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a method for dynamic identification of watersheds during the post-flood season, comprising the following steps: S1. Collect and organize historical data on hydrological factors in the study area, and determine the range of each hydrological factor based on the historical data; denote the lower limit of the historical data range as 0, and the upper limit of the historical data range as... , The number of stations that collected historical data for the study area is used to determine whether the basin has exited the plum rain season based on measured rainfall and atmospheric circulation data. If the basin has exited the plum rain season, it enters the S2 phase; if it has not exited the plum rain season, the basin has not entered the post-flood season. S2, Production Group obedience A uniformly distributed random number combination, let For the first The first in the group The random numbers from each site follow a uniform distribution: ; All random numbers Relatively independent distribution; For each group Water condition factor threshold vector for: ; S3. Select representative basin-type and regional typical flood processes from historical measured data as the test set for each group of hydrological factors; S4. The earliest date within the year when the water levels of all major control stations in the test set determined in S3 are below the warning level and have not subsequently exceeded the warning level is considered the time when there is no flood risk. ; ; in, This represents the time difference value; For the first The group of water-related factors has entered the post-flood season period; This is a period with no flood risk. For the first The time when each station is free from flood risk; S5. Compare the threshold values ​​of each randomly generated hydrological factor in S2 with those of the typical regional flood process in S3 on a day-by-day basis. For each group... and each site Find the earliest date This makes from Initially, the water level was always below or equal to the water condition factor threshold vector. : ; in, For the first Group Sites In time Water-related factors; S6. Compare with S5 to obtain the time difference value. Select the set of hydrological factors with the smallest differences. : .

[0008] Furthermore, in S1, for hydrological factors in the study area, direct hydraulic factors are selected in areas with complete river hydrological station networks; for areas without flow monitoring stations, average precipitation is selected as a substitute or supplementary indicator, and a reasonable numerical range is determined based on the historical data sequence of the selected factors to provide boundary conditions for subsequent simulations.

[0009] Furthermore, in S1, the historical data is required to be more than 30 years to ensure that it can comprehensively cover various hydrological year types such as abundant, normal, and dry years within the basin; Furthermore, whether the plum rain season has ended in S1 is determined by a combination of the average rainfall over the basin area and characteristic atmospheric circulation indicators.

[0010] Furthermore, in S2, production Group obedience Uniformly distributed random number combinations are used for random sampling when there is a lack of effective prior information; however, in areas where hydrological characteristics have been studied in depth, a distribution form that better matches the probability characteristics of hydrological factors is selected based on historical statistical features. To avoid the randomness of Monte Carlo simulation from having a chance effect on the results, the number of random arrays is more than 1000 to ensure the stability of the statistical results.

[0011] Furthermore, in S3, typical flood processes of the basin and the region should be selected based on the principles of flood magnitude, flood hydrograph morphology, and adverse effects on flood control.

[0012] Furthermore, in step 1, the main control stations need to be representative control stations within the basin; the main control stations should cover the main stream control section and the main tributary confluence points that have a significant impact on flood control safety. The warning water level is derived from basic data from various hydrological stations.

[0013] Furthermore, a dynamic identification method for the post-flood season of a river basin is implemented through the following modules: The data collection and plum rain season exit determination module is used to collect historical data of hydrological factors in the study area, determine the range of factor intervals, and determine whether the watershed has exited the plum rain season based on measured rainfall and atmospheric circulation data, thus providing a prerequisite for identification in the post-flood season. The random threshold generation module for hydrological factors is used to generate multiple sets of threshold vectors for hydrological factors through Monte Carlo simulation, providing diverse boundary conditions for identification during the post-flood season. The typical flood test set selection module is used to select representative flood processes to form a test set, providing basic data for threshold validity verification. The flood-free risk time estimation module is used to determine the flood-free risk time for the basin and each station, providing a calibration benchmark for threshold optimization; The post-flood season start time calculation module is used to compare each group of thresholds with the test set and calculate the corresponding watershed post-flood season start time and time difference value for each group. The hydrological factor threshold optimization module is used to screen the optimal hydrological factor thresholds and determine the final conditions for dynamic identification of the basin after the flood season.

[0014] The beneficial effects of this invention are as follows: This method introduces the Monte Carlo random simulation method into the study of flood season stages for the first time, and establishes a dynamic identification technology based on the combination of thresholds of multiple water conditions.

[0015] Compared with traditional static methods based on fixed statistical periods, this invention has three significant advantages: First, it effectively solves the problem of flood season division caused by the complex composition of flood-prone areas through large-scale random sampling, and can adapt to the receding characteristics of multi-source asynchronous floods; second, it proposes a flexibly configurable hydrological factor determination mechanism, which can adopt strict determination of all factors or identification based on key factors, thereby enhancing the applicability of the method under different geographical conditions; third, it innovatively uses the time without flood risk as a calibration benchmark, and achieves a balance between flood control safety and water resource utilization through a flood control safety strategy-oriented optimization mechanism.

[0016] This method provides new technical support for flood control, drought relief, and refined water resource management in large river basins, especially in areas with complex flood compositions. Attached Figure Description

[0017] Figure 1 A flowchart of a method for dynamic identification of watersheds during the post-flood season; Figure 2 This is a comparison chart of boundary conditions for water-related factors and time periods without flood risk. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0019] like Figure 1 As shown, a method for dynamic identification of watersheds during the post-flood season includes the following steps: S1. Collect and organize historical data on hydrological factors in the study area for 73 years, from 1952 to 2024, such as daily average water level, flow rate, rainfall, and atmospheric circulation data. Determine the range of each hydrological factor based on the historical data; denote the lower limit of the historical data range as 0, and the upper limit of the historical data range as... , The number of stations that collected historical data in the study area; if the daily rainfall in the basin is less than 5 mm for 5 consecutive days, and the subtropical high pressure ridge jumps northward to ≥27°N and its intensity increases significantly, then the basin is judged to have exited the plum rain season and S2 is carried out; otherwise, the basin has not entered the post-flood season. in, It is 7; S2, Production Group obedience A uniformly distributed random number combination, let For the first The first in the group The random numbers from each site follow a uniform distribution: ; All random numbers Relatively independent distribution; For each group Water condition factor threshold vector for: ; in, =1,2,…,7; It is 10000; S3. Select representative basin-type and regional typical flood processes from historical measured data as the test set for each group of hydrological factors; Typical flood years are 1954, 1998, and 2020; typical years of normal water levels are 2011, 2013, and 2017; and typical years of low water levels are 1978, 2006, and 2022. S4. The earliest date within the year when the water levels of all major control stations in the test set determined in S3 are below the warning level and have not subsequently exceeded the warning level is considered the time when there is no flood risk. ; ; in, This represents the time difference value; For the first The group of water-related factors has entered the post-flood season period; This is a period with no flood risk. For the first The time when each station is free from flood risk; S5. Compare the threshold values ​​of each randomly generated hydrological factor in S2 with those of the typical regional flood process in S3 on a day-by-day basis. For each group... and each site Find the earliest date This makes from Initially, the water level was always below or equal to the water condition factor threshold vector. : ; in, For the first Group Sites In time Water-related factors; S6. Compare with S5 to obtain the time difference value. Select the set of hydrological factors with the smallest differences. : .

[0020] No. Group water condition factor threshold This is the criterion for determining whether a river basin has entered the post-flood season.

[0021] A comparison chart of the optimized hydrological factor boundary conditions and the time without flood risk is shown below. Figure 2 .

[0022] In S1, for hydrological factors in the study area, direct hydraulic factors are selected in areas with complete river hydrological station networks; for areas without flow monitoring stations, average precipitation is selected as a substitute or supplementary indicator, and a reasonable numerical range is determined based on the historical data sequence of the selected factors to provide boundary conditions for subsequent simulations.

[0023] In S1, historical data is required to be more than 30 years to ensure that it can comprehensively cover various hydrological year types such as abundant, normal, and dry years within the basin; Whether the plum rain season has ended in S1 is determined by a combination of the average rainfall over the basin area and characteristic atmospheric circulation indicators.

[0024] In S2, production Group obedience Uniformly distributed random number combinations are used for random sampling when there is a lack of effective prior information. In areas where hydrological characteristics are studied in depth, a distribution form that better matches the probability characteristics of hydrological factors is selected based on historical statistical features, such as the normal distribution or the Gamma distribution.

[0025] To avoid the randomness of Monte Carlo simulation from having a chance effect on the results, the number of random arrays is more than 1000 to ensure the stability of the statistical results.

[0026] In S3, typical flood processes of the basin and the region should be selected based on the principles of flood magnitude, flood hydrograph morphology, and adverse effects on flood control.

[0027] In step 1, the main control stations need to be representative control stations within the watershed; the main control stations should cover the main stream control section and the main tributary confluence points that have an important impact on flood control safety. The warning water level is derived from basic data from various hydrological stations. It can also be set as a guaranteed water level according to the specific needs of the river basin.

[0028] Select The latest time at which there is no flood risk among the various stations This period can be considered as a time without flood risk in the basin. Alternatively, based on the importance of the stations or the characteristics of the flood-prone area, key stations can be selected. This period is considered to be free of flood risk in the basin. ; Select The latest of all the sites The time T is considered as the start of the post-flood season for the river basin. Alternatively, key stations can be selected based on their importance or the characteristics of the flood-affected area. The time T is considered as the start of the post-flood season for this river basin; The boundary condition that "all hydrological factors (water level, flow rate) in a typical year are less than those of randomly generated hydrological factors" can be flexibly set according to actual application needs: a strict mode can be adopted, requiring all hydrological factors (water level, flow rate, etc.) to be lower than the corresponding threshold; or a critical factor mode can be adopted, that is, only some critical hydrological factors that play a decisive role in flood control risk need to meet the threshold condition. The minimum difference criterion can be flexibly adjusted according to the flood control needs of the watershed: the preferred criterion is... The one with the smallest absolute difference from Y is selected when multiple sets of threshold differences have the same absolute value. The selection is based on the watershed flood control safety strategy: for watersheds with high flood control requirements, the corresponding later threshold is selected. The threshold condition; for watersheds with urgent water resource utilization needs, the corresponding earlier threshold condition can be selected. The corresponding hydrological factors serve as criteria for determining whether the basin has entered the post-flood season.

[0029] Furthermore, a dynamic identification method for the post-flood season of a river basin is implemented through the following modules: The data collection and plum rain season exit determination module is used to collect historical data of hydrological factors in the study area, determine the range of factor intervals, and determine whether the watershed has exited the plum rain season based on measured rainfall and atmospheric circulation data, thus providing a prerequisite for identification in the post-flood season. The random threshold generation module for hydrological factors is used to generate multiple sets of threshold vectors for hydrological factors through Monte Carlo simulation, providing diverse boundary conditions for identification during the post-flood season. The typical flood test set selection module is used to select representative flood processes to form a test set, providing basic data for threshold validity verification. The flood-free risk time estimation module is used to determine the flood-free risk time for the basin and each station, providing a calibration benchmark for threshold optimization; The post-flood season start time calculation module is used to compare each group of thresholds with the test set and calculate the corresponding watershed post-flood season start time and time difference value for each group. The hydrological factor threshold optimization module is used to screen the optimal hydrological factor thresholds and determine the final conditions for dynamic identification of the basin after the flood season.

[0030] The embodiments described above are merely illustrative of implementation methods of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be defined by the appended claims.

Claims

1. A method for dynamic identification of watersheds during the post-flood season, characterized in that, Includes the following steps: S1. Collect and organize historical data on hydrological factors in the study area, and determine the range of each hydrological factor based on the historical data; denote the lower limit of the historical data range as 0, and the upper limit of the historical data range as... , The number of stations that collected historical data for the study area is used to determine whether the basin has exited the plum rain season based on measured rainfall and atmospheric circulation data. If the basin has exited the plum rain season, it enters the S2 phase; if it has not exited the plum rain season, the basin has not entered the post-flood season. S2, Production Group obedience A uniformly distributed random number combination, let For the first The first in the group The random numbers from each site follow a uniform distribution: ； All random numbers Relatively independent distribution; For each group Water condition factor threshold vector for: ； S3. Select representative basin-type and regional typical flood processes from historical measured data as the test set for each group of hydrological factors; S4. The earliest date within the year when the water levels of all major control stations in the test set determined in S3 are below the warning level and have not subsequently exceeded the warning level is considered the time when there is no flood risk. ； ； in, This represents the time difference value; For the first The group of water-related factors has entered the post-flood season period; This is a period with no flood risk. For the first The time when each station is free from flood risk; S5. Compare the threshold values ​​of each randomly generated hydrological factor in S2 with those of the typical regional flood process in S3 on a day-by-day basis. For each group... and each site Find the earliest date This makes from Initially, the water level was always below or equal to the water condition factor threshold vector. : ； in, For the first Group Sites In time Water-related factors; S6. Compare with S5 to obtain the time difference value. Select the set of hydrological factors with the smallest differences. : 。 2. The method for dynamic identification of watersheds during the post-flood season according to claim 1, characterized in that: In S1, for hydrological factors in the study area, direct hydraulic factors are selected in areas with complete river hydrological station networks; for areas without flow monitoring stations, average precipitation is selected as a substitute or supplementary indicator, and a reasonable numerical range is determined based on the historical data sequence of the selected factors to provide boundary conditions for subsequent simulations.

3. The method for dynamic identification of watersheds during the post-flood season according to claim 1, characterized in that: In S1, historical data is required to be over 30 years to ensure a relatively complete coverage of various hydrological year types, including abundant, normal, and dry years, within the basin.

4. The method for dynamic identification of watersheds during the post-flood season according to claim 1, characterized in that: Whether the plum rain season has ended in S1 is determined by a combination of the average rainfall over the basin area and characteristic atmospheric circulation indicators.

5. The method for dynamic identification of watersheds during the post-flood season according to claim 1, characterized in that: In S2, production Group obedience Uniformly distributed random number combinations are used for random sampling when there is a lack of effective prior information; however, in areas where hydrological characteristics have been studied in depth, a distribution form that better matches the probability characteristics of hydrological factors is selected based on historical statistical features. To avoid the randomness of Monte Carlo simulation from having a chance effect on the results, the number of random arrays is more than 1000 to ensure the stability of the statistical results.

6. The method for dynamic identification of watersheds during the post-flood season according to claim 1, characterized in that: In S3, typical flood processes of the basin and the region should be selected based on the principles of flood magnitude, flood hydrograph morphology, and adverse effects on flood control.

7. The method for dynamic identification of watersheds during the post-flood season according to claim 1, characterized in that: In step 1, the main control stations need to be representative control stations within the watershed; the main control stations should cover the main stream control section and the main tributary confluence points that have an important impact on flood control safety. The warning water level is derived from basic data from various hydrological stations.

8. A method for dynamic identification of watersheds during the post-flood season, characterized in that: This can be achieved through the following modules: The data collection and plum rain season exit determination module is used to collect historical data of hydrological factors in the study area, determine the range of factor intervals, and determine whether the watershed has exited the plum rain season based on measured rainfall and atmospheric circulation data, thus providing a prerequisite for identification in the post-flood season. The random threshold generation module for hydrological factors is used to generate multiple sets of threshold vectors for hydrological factors through Monte Carlo simulation, providing diverse boundary conditions for identification during the post-flood season. The typical flood test set selection module is used to select representative flood processes to form a test set, providing basic data for threshold validity verification. The flood-free risk time estimation module is used to determine the flood-free risk time for the basin and each station, providing a calibration benchmark for threshold optimization; The post-flood season start time calculation module is used to compare each group of thresholds with the test set and calculate the corresponding watershed post-flood season start time and time difference value for each group. The hydrological factor threshold optimization module is used to screen the optimal hydrological factor thresholds and determine the final conditions for dynamic identification of the basin after the flood season.

Images