Historical flood recurrence period determination method based on multi-section flood correlation

Abstract

The invention discloses a historical flood recurrence period determination method based on multi-section flood correlation, and the method comprises the steps: recognizing a hydrological homogeneous region (target river reach) with the flood law having high spatial consistency in a drainage basin, and selecting a section with the most complete data as a reference for precise calibration; and the calibrated recurrence period is scientifically popularized to the whole homogeneous region based on the physical consistency principle. According to the method, the target river reach is screened through the correlation between the multi-section flood and the area, the problem of recurrence period discretization caused by section isolated analysis in a traditional method is avoided, a consistent reference is provided for basin flood control planning, the recurrence period error is controlled within 3.2% through initial determination of the recurrence period range and then accurate calibration through the line adaptation method, subjective judgment is avoided, and the accuracy of flood control is improved. The method can be directly applied to similar mountainous area drainage basins, does not depend on long-sequence observation data, and is remarkable in practicability.

Description

Technical Field

[0001] This invention belongs to the field of hydrological analysis technology for water conservancy projects, and relates to a method for determining the return period of historical floods based on the correlation of floods across multiple cross sections. Background Technology

[0002] Historical flood recurrence intervals are the core basis for formulating flood control standards for water conservancy projects. In vast mountainous watersheds, hydrological stations are generally scarce and observation periods are short, far from meeting the sample length requirements for frequency analysis of floods with recurrence intervals of "once in a century" or even longer. Therefore, relying on historical flood surveys (including literature review, field surveys, and flood trace surveys) to extend flood sequences has become an indispensable technical means.

[0003] However, current methods for determining the recurrence interval of historical floods have significant drawbacks: 1. Subjectivity and Inconsistency in Verification Periods: The criteria for determining the year of the "first credible historical flood" heavily rely on the subjective experience and judgment of analysts. In practice, different sections within the same watershed often use vastly different starting points for verification periods (e.g., some use 1949 as the boundary, while others can be traced back to 19th-century documents). This directly leads to discrepancies in the calculated return periods for each section, with deviations reaching 30% to 50%, thus disrupting the overall coordination between watershed flood control planning and engineering design.

[0004] 2. Uncertainty of survey information: Except for a few sections with detailed documentary evidence, most sections rely on oral accounts of historical floods, which leads to memory bias of flood years and errors in flood mark identification, resulting in significant errors in the estimated peak flow.

[0005] 3. Lack of Spatial Correlation Verification: Traditional methods typically treat each cross-section as an independent entity for isolated analysis, neglecting the inherent spatial and physical correlations of flood generation, confluence, and evolution within the watershed. Flood responses in mountainous watersheds are strongly controlled by underlying surface conditions such as topography, vegetation, and river network structure. Within the same hydrological response unit, flood characteristics (such as peak modulus) should exhibit good regularity. Ignoring this correlation results in isolated return period results obtained independently from each cross-section, lacking support and mutual verification from the watershed's hydrophysical mechanisms.

[0006] The aforementioned problems result in insufficient reliability and consistency of existing historical flood analysis results for mountainous watersheds, potentially leading to inadequate flood control capacity in water conservancy projects or unnecessary investment waste. Therefore, there is an urgent need for an innovative method that can integrate watershed spatial information, be based on the physical laws of flood formation, and achieve coordinated and unified determination of return periods across multiple cross sections. Summary of the Invention

[0007] The purpose of this invention is to provide a method for determining the return period of historical floods based on the correlation of floods across multiple cross sections, which solves the problems in the prior art of subjective verification of historical flood return periods, discrete return period results caused by isolated cross section analysis, and lack of physical basis.

[0008] The technical solution adopted in this invention is a method for determining the return period of historical floods based on the correlation of floods across multiple sections. This method is applicable to mountainous watersheds with sparse hydrological stations and short observation sequences. The specific implementation steps are as follows: Step 1: Collect basic data from multiple spatially continuous survey sections within the target mountainous watershed and verify the reliability of the data; Step 2: Identify the target river segment based on the correlation between flood and area; Step 3: Select a benchmark section within the target river section and preliminarily determine the return period range; Step 4: Calibrate the return period within the range based on frequency analysis; Step 5: Based on spatial consistency, unify the calibration return period of the benchmark section to the return period of all survey sections within the target river section.

[0009] The invention is further characterized by: Step 1 requires that the survey sections include at least one hydrological station with a long series of measured annual maximum flood peak flow data and two sections with reliable historical flood survey information; the historical flood survey information includes flood peak flow and return period; the basic data includes the catchment area of ​​each section and information on the first major historical flood; the historical flood survey information includes the year of occurrence of the historical flood, flood peak flow and return period.

[0010] The data reliability verification in step 1 is carried out using a multi-level verification method, including source tracing verification, logical consistency verification of upstream and downstream flood events, and cross-verification of literature from different sources. Specifically, source tracing verification involves checking the original source and rationality of the key hydraulic parameters used to estimate peak flow, including flood mark location, channel roughness, and water surface gradient. Logical consistency verification involves verifying whether the peak flow of the same basin-wide historical flood event at different upstream and downstream sections follows the hydrophysical law of increasing with the increase of catchment area. Cross-verification of literature involves comparing the year and disaster description of the same historical flood event in literature from different sources, and correcting it based on the authoritative record of the most recent section in terms of time and space.

[0011] The specific steps of step 2 are as follows: Step 2-1: Select the peak flow and catchment area of ​​the first major historical flood at each cross section as the analysis sample; Step 2-2: Plot the analysis sample points on the double logarithmic coordinate graph and fit the power function curve of the flow-area relationship model; Steps 2-3: Screening for goodness of fit R2 The set of continuous cross sections that are not lower than the preset threshold is the target river section.

[0012] The power function of the flow-area relationship model is: (1); In formula (1), This is the regional comprehensive coefficient. Area index; Goodness of fit R 2 The preset threshold is 0.95.

[0013] Step 3 specifically involves: Step 3-1: Select the cross section with the longest historical flood record sequence and the most complete and reliable information within the target river section as the benchmark cross section; Step 3-2: Based on the historical flood sequence records of this benchmark section, determine the first major historical flood event during the investigation period; Step 3-3: By comprehensively utilizing the historical ranking method and the spatiotemporal response relationship of major flood events in the upper and lower reaches of the basin, the recurrence period of this flood is cross-validated and its range is defined, providing a preliminary estimated range for the recurrence period. .

[0014] Step 4 is as follows: Step 4-1: Construct the annual maximum flood peak flow sequence for the benchmark section. The annual maximum flood peak flow sequence includes historical flood survey data and measured data. Step 4-2: Use P-III type curves to perform frequency analysis and linear fitting on the annual maximum flood peak flow sequence; Step 4-3: When the goodness of fit meets the requirements, the precise statistical return period corresponding to the first major historical flood can be read from the fitted frequency curve. .

[0015] During the line fitting process, the coefficient of variation is adjusted while keeping the key parameters reasonable. and skewness coefficient ,Keep =3.00.

[0016] The principle of alignment is to control the relative deviation between the historical flood data and the theoretical frequency curve within a set range.

[0017] The relative deviation is no greater than 5%; the goodness of fit R in step 4-3 2 The requirement is not less than 0.95.

[0018] The beneficial effects of this invention are: This invention selects target river sections by correlating flood data with area across multiple cross sections, avoiding the return period discrepancy problem caused by "isolated cross section analysis" in traditional methods. It provides a consistent benchmark for watershed flood control planning. After initially determining the return period range, it is then precisely calibrated using the fitting method, keeping the return period error within 3.2%. This avoids subjective judgment and can be directly applied to similar mountainous watersheds without relying on long-sequence observation data, demonstrating significant practicality. Attached Figure Description

[0019] Figure 1 This is a diagram showing the relationship between the largest historical flood and the catchment area at multiple cross-sections of the main stream of the Jialing River in Embodiment 1 of the present invention. Detailed Implementation

[0020] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0021] This invention provides a method for determining the historical flood return period based on multi-section flood correlation. It identifies a hydrologically homogeneous area (target river segment) within a watershed with highly spatially consistent flood patterns, selects the section with the most complete data as a benchmark for precise calibration, and then scientifically extends the calibrated return period to the entire homogeneous area based on the principle of physical consistency. The specific implementation steps are as follows: Step 1: Collect basic data from multiple spatially continuous survey sections within the target mountainous watershed and verify the reliability of the data; The survey sections include at least one hydrological station with a long series of measured annual maximum flood peak flow data and two sections with reliable historical flood survey information; the catchment area of ​​the survey sections is calculated using a 1:50,000 topographic map (accuracy ±2%).

[0022] Basic data includes the catchment area of ​​each section and information on the largest historical flood (i.e., the largest flood event ranked first during the survey period); historical flood survey information includes the year of occurrence of historical floods, peak flow, recurrence period, data reliability, and sources of supporting documents (such as local chronicles and hydrological archives).

[0023] Data reliability verification is carried out using a multi-level verification method, including data source tracing verification, logical consistency verification of upstream and downstream flood events, and cross-verification of documents from different sources. Data tracing and verification: Verify the original source and rationality of the key hydraulic parameters used in the flood peak flow estimation, including flood mark location, river roughness, and water surface gradient; Logical consistency verification: Verify whether the peak flow of the same basin-wide historical flood event at different upstream and downstream sections follows the hydrophysical law that increases with the increase of the catchment area; Cross-reference of documents: Compare the year and disaster description of the same historical flood event in documents from different sources (such as county annals, flood survey compilations, and scientific research reports), and make corrections based on the authoritative record that is closest to the cross section in time and space.

[0024] Step 2: Identify the target river segment based on the correlation between flood and area; The target river section is a hydrologically homogeneous area; this step aims to identify spatial units with uniform flood patterns from a physical mechanism perspective. The specific steps are as follows: Step 2-1: Select the peak flow and catchment area of ​​the first major historical flood at each cross section as the analysis sample; during the data selection process, remove abnormal data caused by large-scale tributary inflows and human activities. Step 2-2: Plot the analysis sample points on the double logarithmic coordinate graph and fit the power function curve of the flow-area relationship model; In regions with similar climate and underlying surface conditions, peak flood flow With water catchment area There is usually an exponential function relationship, therefore the exponential function is: (1); In formula (1), This is the regional comprehensive coefficient. This is the area index.

[0025] Steps 2-3: Screening for goodness of fit R 2 The set of continuous cross sections not lower than a preset threshold is the target river section; the preset threshold is 0.95.

[0026] Goodness of fit R 2 It is a core indicator for quantifying spatial consistency. When a series of When the goodness of fit of the data reaches or exceeds a preset high threshold (e.g., 0.95, 0.98, or 0.99), these cross sections are determined to be controlled by similar hydrophysical processes, and their flood formation and evolution patterns exhibit a high degree of spatial consistency. Therefore, they are collectively designated as the same "hydrophysical homogeneous zone (target river section)". This high threshold standard ensures the homogeneity of flood characteristics within the unit, laying a solid physical foundation for subsequent unified extrapolation of return periods.

[0027] Step 3: Select a benchmark section within the target river section and preliminarily determine the return period range; Step 3-1: Select the cross section with the longest historical flood record sequence and the most complete and reliable information within the target river section as the benchmark cross section; Step 3-2: Based on the historical flood sequence records of this benchmark section, determine the first major historical flood event during the investigation period; Step 3-3: By comprehensively utilizing the historical ranking method and the spatiotemporal response relationship of major flood events in the upper and lower reaches of the basin, the recurrence period of this flood is cross-validated and its range is defined, providing a preliminary estimated range for the recurrence period. .

[0028] Step 4: Calibrate the return period within the range based on frequency analysis; Step 4-1: Construct the annual maximum flood peak flow sequence for the benchmark section. The annual maximum flood peak flow sequence includes historical flood survey data and measured data. Step 4-2: Use P-III type curves to perform frequency analysis and linear fitting on the annual maximum flood peak flow sequence; During the line fitting process, the coefficient of variation is adjusted while keeping key parameters such as the mean reasonable. and skewness coefficient ,Keep =3.00. The fitting principle focuses on fitting the upper and middle parts of the empirical data points, especially ensuring that the relative deviation between the data points of the first major historical flood and the theoretical frequency curve is controlled within the set range, with a relative deviation of no more than 5%.

[0029] Step 4-3: When the goodness of fit meets the requirements, the precise statistical return period T corresponding to the first major historical flood can be read from the fitted frequency curve. c This step transforms the preliminary judgment based on literature into a precise inference based on mathematical statistics.

[0030] Step 5: Based on spatial consistency, unify the calibration return period of the benchmark section to the return period of all survey sections within the target river section; The return period T for calibration of the reference section determined in step 4. c This directly assigns all other survey sections within the target river section the recurrence period of their respective first major historical flood.

[0031] Based on the high spatial consistency of flood patterns within the target river section, it can be determined from a hydrological perspective that, although the years of occurrence and peak flow values ​​of the first major historical flood events occurring at different cross-sections within this hydrologically homogeneous area may differ, the probability of extreme flood occurrence (i.e., return period) represented by them at their respective locations is the same or extremely close. This allows for the efficient and scientific spatial unification and standardization of the historical flood return period throughout the entire hydrologically homogeneous area.

[0032] The historical flood return period determination method based on multi-section flood correlation of this invention is applicable to mountainous watersheds with sparse hydrological stations and short observation sequences (such as the main stream of the Jialing River above Shaanxi Province and similar Qinling-Bashan mountainous watersheds). It provides unified and accurate historical flood return period benchmark data for watershed flood control planning and water conservancy project (such as dikes and reservoirs) design, supporting decisions on balancing project safety and economy; it has the following advantages: 1. Fundamentally improve the uniformity and coordination of results: By identifying hydrological homogeneous zones through objective mathematical standards (high goodness of fit), the subjective practice of manually defining verification periods has been completely replaced. This ensures that the return period results of all sections within the same physical unit are naturally unified and mutually coordinated, providing a solid and consistent foundation for the overall flood control planning and engineering design of the basin.

[0033] 2. Significantly improves the objectivity and scientific rigor of the determination process: the entire process is based on data relationships ( With correlation, statistical models (P-III curves) and standardized criteria (fitting requirements) as the core drivers, the process minimizes subjective judgment errors caused by individual differences, and transforms the process of determining the return period from "experience-driven" to a standardized technical process driven by "data and models".

[0034] 3. Supported by a solid physical mechanism: The method creatively uses the core hydrological principle of "spatial correlation of floods" as the cornerstone of the entire method, which makes the extension of the return period from a single point to the watershed area have clear physical significance and logical necessity, and the conclusions are more reliable.

[0035] 4. Highly applicable to areas with scarce hydrological data: This method does not rely on long-sequence measured flow data, but fully explores and integrates multi-source heterogeneous information such as historical surveys, literature records, and existing measured data. It is particularly suitable for promotion and application in mountainous watersheds with scarce hydrological data, and effectively solves the problem of determining key hydrological parameters in engineering design in these areas.

[0036] The technical solution of this invention has broad applicability and can be universally applied to any mountainous river basin with the following characteristics: 1. Sparse hydrological stations, with flood control design mainly relying on historical flood survey data; 2. The existence of continuous river sections with relatively uniform underlying surface conditions such as climate, topography, and vegetation within the basin; 3. The availability of a certain number of historical documents (such as county annals, hydrological archives, flood stone inscriptions, etc.) as verification basis. In application, the return period can be uniformly determined simply by strictly following the standard steps described in the claims and above-mentioned content of this invention, based on the specific data of the new basin.

[0037] Example 1 In this embodiment, the return period of the Jialing River main stream above Shaanxi Province is determined using a historical flood return period determination method based on multi-section flood correlation. The specific implementation steps are as follows: Step 1: Collect basic data from multiple spatially continuous survey sections within the target mountainous watershed and verify the reliability of the data; Within this watershed, nine spatially continuous survey sections were selected: Huangniupu, Honghuapu, Fengzhou, Ciba, Tanjiazhuang, Baishuijiang Town, Lueyang, Maolongba, and Xindianzi. Historical flood data are sourced from sources such as "Shaanxi Province Flood Survey Data", "Feng County Annals", and "Lueyang County Annals". The reliability of the data was verified through a multi-level verification method, including source tracing verification, logical consistency verification of upstream and downstream flood events, and cross-verification of literature from different sources, to ensure the reliability of the key flood data used for analysis.

[0038] Step 2: Identify the target river segment based on the correlation between flood and area; The peak flow of the first major historical flood at each of the above nine cross sections. With water catchment area Plotting points on a logarithmic coordinate graph, such as Figure 1 As shown, the identification process and screening criteria of the hydrological homogeneous zone (target river section) are intuitively demonstrated. The figure clearly shows that the data of 6 continuous cross sections are closely distributed on both sides of the fitted line, forming a highly correlated hydrological homogeneous zone (target river section); the data of the other 3 cross sections are discrete due to changes in hydrological conditions.

[0039] Therefore, the fitting results show that the data points at the six continuous cross-sections from Huangniupu to Baishuijiang Town exhibit extremely high linear correlation. The fitting formula is as follows: Goodness of fit R 2 =0.9988, and a high goodness-of-fit threshold R is set. 2 If the value is 0.95, then these six cross-sections fully meet this condition and are therefore identified and designated as the target river section. The three cross-sections downstream of Baishuijiang Town—Lueyang, Maolongba, and Xindianzi—are significantly affected by the inflow of large tributaries such as the Xihan River, resulting in significant changes in the hydrological situation of the basin. Their data points deviate significantly from the trend line of the six upstream points, exhibiting low goodness of fit; therefore, they are not included in this hydrologically homogeneous area (target river section).

[0040] Step 3: Select a benchmark section within the target river section and preliminarily determine the return period range; Within the established hydrologically homogeneous area (6 cross sections), Fengzhou Station has the most continuous and detailed county-level records of flood disasters since 1808, and was therefore selected as the benchmark cross section. According to the *Feng County Gazetteer*, the 1981 flood was the largest since 1808, with a preliminary estimated recurrence interval of approximately 220 years. Combined with the catastrophic flood of 1857 at the downstream Lueyang station, clearly recorded in the *Lueyang County Gazetteer*, and based on flood characteristics and rainfall analysis, it can be inferred that this flood was also a maximum event at the Fengzhou station. If 1857 is taken as the new starting point for verification at the Fengzhou station, then the recurrence interval of the 1981 flood is approximately 170 years. Based on both pieces of evidence, the recurrence period of the first major historical flood at Fengzhou Station in 1981 is preliminarily set at 170 to 220 years.

[0041] Step 4: Calibrate the return period within the range based on frequency analysis; Construct a sequence of the annual maximum flood peak flow at Fengzhou Station, and incorporate the historical flood data from 1981; Frequency fitting was performed using P-III type curves; the coefficient of variation was adjusted. and skewness coefficient This process ensured the best fit between empirical data and the theoretical curve, and guaranteed that the relative deviation between the 1981 flood data and the curve was less than 5%. Ultimately, the 1981 flood at Fengzhou station (1640m) was determined. 3 The exact return period corresponding to / s is 220 years.

[0042] Step 5: Based on spatial consistency, unify the calibration return period of the benchmark section to the return period of all survey sections within the target river section; Based on the extremely high flood-area spatial correlation (Ri) within the hydrologically homogeneous zone (target river segment) verified in step 2. 2 =0.9988), indicating that the flood patterns within this river section are highly consistent. Therefore, the precise 220-year return period of Fengzhou Station is directly assigned to the other 5 cross sections within this hydrologically homogeneous area (target river section).

[0043] The final unified results show that the return periods of the first major historical floods at Huangniupu (1949 flood), Honghuapu (1981 flood), Ciba (1981 flood), Tanjiazhuang (1981 flood), and Baishuijiang Town (1898 flood) are all 220 years. Applying the method of this invention, the previously discrete (77–220 years) return periods within the hydrologically homogeneous area (target river section) were successfully unified into a consistent 220 years. The deviation between the flow calculated using the unified fitting formula in step 2 and the actual surveyed flow at each cross-section is between 0.2% and 3.1%, further verifying the rationality and reliability of the unified return period results from another perspective.

[0044] Example 2 The method for determining the historical flood return period based on multi-section flood correlation in this embodiment is implemented according to the following steps: Step 1: Collect basic data from multiple spatially continuous survey sections within the target mountainous watershed and verify the reliability of the data; Step 2: Identify the target river segment based on the correlation between flood and area; Step 3: Select a benchmark section within the target river section and preliminarily determine the return period range; Step 4: Calibrate the return period within the range based on frequency analysis; Step 5: Based on spatial consistency, unify the calibration return period of the benchmark section to the return period of all survey sections within the target river section.

[0045] Example 3 The method for determining the historical flood return period based on multi-section flood correlation in this embodiment is implemented according to the following steps: Step 1: Collect basic data from multiple spatially continuous survey sections within the target mountainous watershed and verify the reliability of the data; The survey sections should include at least one hydrological station with a long series of measured annual maximum flood peak flow data and two sections with reliable historical flood survey information; the historical flood survey information includes flood peak flow and return period; the basic data include the catchment area of ​​each section and information on the first major historical flood; the historical flood survey information includes the year of occurrence of the historical flood, flood peak flow and return period.

[0046] Data reliability verification is carried out using a multi-level verification method, including data source tracing verification, logical consistency verification of upstream and downstream flood events, and cross-verification of documents from different sources. The data tracing and verification specifically involves: verifying the original source and rationality of the key hydraulic parameters used in the calculation of flood peak flow, including the location of flood marks, river roughness, and water surface gradient; The logical consistency verification specifically involves verifying whether the peak flow of the same basin-wide historical flood event at different upstream and downstream sections follows the hydrophysical law that increases with the increase of the catchment area. Cross-verification of literature specifically involves comparing the year and disaster descriptions of the same historical flood event in literature from different sources, and correcting the record based on the authoritative record that is closest to the cross section in time and space.

[0047] Step 2: Identify the target river segment based on the correlation between flood and area.

[0048] Step 3: Select a benchmark section within the target river section and preliminarily determine the recurrence period range.

[0049] Step 4: Calibrate the return period within the return period range based on frequency analysis.

[0050] Step 5: Based on spatial consistency, unify the calibration return period of the benchmark section to the return period of all survey sections within the target river section.

[0051] Example 4 The method for determining the historical flood return period based on multi-section flood correlation in this embodiment is implemented according to the following steps: Step 1: Collect basic data from multiple spatially continuous survey sections within the target mountainous watershed and verify the reliability of the data; The survey sections should include at least one hydrological station with a long series of measured annual maximum flood peak flow data and two sections with reliable historical flood survey information; the historical flood survey information includes flood peak flow and return period; the basic data include the catchment area of ​​each section and information on the first major historical flood; the historical flood survey information includes the year of occurrence of the historical flood, flood peak flow and return period.

[0052] Data reliability verification is carried out using a multi-level verification method, including data source tracing verification, logical consistency verification of upstream and downstream flood events, and cross-verification of documents from different sources. The data tracing and verification specifically involves: verifying the original source and rationality of the key hydraulic parameters used in the calculation of flood peak flow, including the location of flood marks, river roughness, and water surface gradient; The logical consistency verification specifically involves verifying whether the peak flow of the same basin-wide historical flood event at different upstream and downstream sections follows the hydrophysical law that increases with the increase of the catchment area. Cross-verification of literature specifically involves comparing the year and disaster descriptions of the same historical flood event in literature from different sources, and correcting the record based on the authoritative record that is closest to the cross section in time and space.

[0053] Step 2: Identify the target river segment based on the correlation between flood and area; The specific steps are as follows: Step 2-1: Select the peak flow and catchment area of ​​the first major historical flood at each cross section as the analysis sample; Step 2-2: Plot the analysis sample points on the double logarithmic coordinate graph and fit the power function curve of the flow-area relationship model; The power function of the flow-area relationship model is: (1); In formula (1), This is the regional comprehensive coefficient. Area index; Steps 2-3: Screening for goodness of fit R 2 The set of continuous cross-sections with a value not lower than a preset threshold is the target river segment; Goodness of fit R 2 The preset threshold is 0.95.

[0054] Step 3: Select a benchmark section within the target river section and preliminarily determine the recurrence period range.

[0055] Step 4: Calibrate the return period within the return period range based on frequency analysis.

[0056] Step 5: Based on spatial consistency, unify the calibration return period of the benchmark section to the return period of all survey sections within the target river section.

[0057] Example 5 The method for determining the historical flood return period based on multi-section flood correlation in this embodiment is implemented according to the following steps: Step 1: Collect basic data from multiple spatially continuous survey sections within the target mountainous watershed and verify the reliability of the data; The survey sections should include at least one hydrological station with a long series of measured annual maximum flood peak flow data and two sections with reliable historical flood survey information; the historical flood survey information includes flood peak flow and return period; the basic data include the catchment area of ​​each section and information on the first major historical flood; the historical flood survey information includes the year of occurrence of the historical flood, flood peak flow and return period.

[0058] Step 2: Identify the target river segment based on the correlation between flood and area; The specific steps are as follows: Step 2-1: Select the peak flow and catchment area of ​​the first major historical flood at each cross section as the analysis sample; Step 2-2: Plot the analysis sample points on the double logarithmic coordinate graph and fit the power function curve of the flow-area relationship model; The power function of the flow-area relationship model is: (1); In formula (1), This is the regional comprehensive coefficient. Area index; Steps 2-3: Screening for goodness of fit R 2 The target river segment is a set of continuous cross-sections that are not lower than a preset threshold; the goodness of fit R 2 The preset threshold is 0.95.

[0059] Step 3: Select a benchmark section within the target river section and preliminarily determine the return period range; Specifically: Step 3-1: Select the cross section with the longest historical flood record sequence and the most complete and reliable information within the target river section as the benchmark cross section; Step 3-2: Based on the historical flood sequence records of this benchmark section, determine the first major historical flood event during the investigation period; Step 3-3: By comprehensively utilizing the historical ranking method and the spatiotemporal response relationship of major flood events in the upper and lower reaches of the basin, the recurrence period of this flood is cross-validated and its range is defined, providing a preliminary estimated range for the recurrence period. .

[0060] Step 4: Calibrate the return period within the return period range based on frequency analysis.

[0061] Step 5: Based on spatial consistency, unify the calibration return period of the benchmark section to the return period of all survey sections within the target river section.

[0062] Example 6 The method for determining the historical flood return period based on multi-section flood correlation in this embodiment is implemented according to the following steps: Step 1: Collect basic data from multiple spatially continuous survey sections within the target mountainous watershed and verify the reliability of the data; The survey sections should include at least one hydrological station with a long series of measured annual maximum flood peak flow data and two sections with reliable historical flood survey information; the historical flood survey information includes flood peak flow and return period; the basic data include the catchment area of ​​each section and information on the first major historical flood; the historical flood survey information includes the year of occurrence of the historical flood, flood peak flow and return period.

[0063] Step 2: Identify the target river segment based on the correlation between flood and area; The specific steps are as follows: Step 2-1: Select the peak flow and catchment area of ​​the first major historical flood at each cross section as the analysis sample; Step 2-2: Plot the analysis sample points on the double logarithmic coordinate graph and fit the power function curve of the flow-area relationship model; The power function of the flow-area relationship model is: (1); In formula (1), This is the regional comprehensive coefficient. Area index; Steps 2-3: Screening for goodness of fit R 2 The target river segment is a set of continuous cross-sections that are not lower than a preset threshold; the goodness of fit R 2 The preset threshold is 0.95.

[0064] Step 3: Select a benchmark section within the target river section and preliminarily determine the return period range; Specifically: Step 3-1: Select the cross section with the longest historical flood record sequence and the most complete and reliable information within the target river section as the benchmark cross section; Step 3-2: Based on the historical flood sequence records of this benchmark section, determine the first major historical flood event during the investigation period; Step 3-3: By comprehensively utilizing the historical ranking method and the spatiotemporal response relationship of major flood events in the upper and lower reaches of the basin, the recurrence period of this flood is cross-validated and its range is defined, providing a preliminary estimated range for the recurrence period. .

[0065] Step 4: Calibrate the return period within the range based on frequency analysis; Specifically: Step 4-1: Construct the annual maximum flood peak flow sequence for the benchmark section. The annual maximum flood peak flow sequence includes historical flood survey data and measured data. Step 4-2: Use P-III type curves to perform frequency analysis and linear fitting on the annual maximum flood peak flow sequence; During the line fitting process, the coefficient of variation is adjusted while keeping key parameters such as the mean reasonable. and skewness coefficient ,Keep =3.00; The fitting principle focuses on fitting the data of major floods in the upper and middle parts of the empirical data to ensure that the relative deviation between the data of the first major historical flood and the theoretical frequency curve is no more than 5%.

[0066] Step 4-3: When the goodness of fit is not less than 0.95, the precise statistical return period corresponding to the first major historical flood can be read from the fitted frequency curve. .

[0067] Step 5: Based on spatial consistency, unify the calibration return period of the benchmark section to the return period of all survey sections within the target river section.

Claims

1. A method for determining the return period of historical floods based on multi-section flood correlation, applicable to mountainous watersheds with sparse hydrological stations and short observation sequences, characterized in that... The specific steps are as follows: Step 1: Collect basic data from multiple spatially continuous survey sections within the target mountainous watershed and verify the reliability of the data; Step 2: Identify the target river segment based on the correlation between flood and area; Step 3: Select a benchmark section within the target river section and preliminarily determine the return period range; Step 4: Calibrate the return period within the return period range based on frequency analysis; Step 5: Based on spatial consistency, unify the calibration return period of the benchmark section to the return period of all survey sections within the target river section.

2. The method for determining the return period of historical floods based on multi-section flood correlation according to claim 1, characterized in that, In step 1, the survey sections include at least one hydrological station with a long series of measured annual maximum flood peak flow data and two sections with reliable historical flood survey information; the historical flood survey information includes flood peak flow and return period. Basic data includes the catchment area of ​​each cross section and information on the first major historical flood; historical flood survey information includes the year of occurrence of historical floods, peak flow, and recurrence period.

3. The method for determining the return period of historical floods based on multi-section flood correlation according to claim 1, characterized in that, The data reliability verification in step 1 is carried out using a multi-level verification method, including data source tracing verification, logical consistency verification of upstream and downstream flood events, and cross-verification of documents from different sources. The data tracing and verification specifically involves: verifying the original source and rationality of the key hydraulic parameters used in the calculation of flood peak flow, including the location of flood marks, river roughness, and water surface gradient; The logical consistency verification specifically involves verifying whether the peak flow of the same basin-wide historical flood event at different upstream and downstream sections follows the hydrophysical law that increases with the increase of the catchment area. Cross-verification of literature specifically involves comparing the year and disaster descriptions of the same historical flood event from different sources, and correcting them based on the authoritative record of the most recent cross section in time and space.

4. The method for determining the return period of historical floods based on multi-section flood correlation according to claim 1, characterized in that, The specific steps of step 2 are as follows: Step 2-1: Select the peak discharge and catchment area of ​​the first major historical flood at each cross section as the analysis sample; Step 2-2: Plot the analysis sample points on the double logarithmic coordinate graph and fit the power function curve of the flow-area relationship model; Steps 2-3: Screening for goodness-of-fit R 2 The set of continuous cross sections that are not lower than the preset threshold is the target river section.

5. The method for determining the return period of historical floods based on multi-section flood correlation according to claim 4, characterized in that, The power function of the flow-area relationship model is: （1）； In formula (1), This is the regional comprehensive coefficient. Area index; The goodness of fit R in steps 2-3 2 The preset threshold is 0.

95.

6. The method for determining the return period of historical floods based on multi-section flood correlation according to claim 1, characterized in that, Step 3 specifically involves: Step 3-1: Select the cross section with the longest historical flood record sequence and the most complete and reliable information within the target river section as the benchmark cross section; Step 3-2: Based on the historical flood sequence records of this benchmark section, determine the first major historical flood event during the investigation period; Step 3-3: By comprehensively utilizing the historical ranking method and the spatiotemporal response relationship of major flood events in the upper and lower reaches of the basin, the recurrence period of this flood is cross-validated and its range is defined, providing a preliminary estimated range for the recurrence period. .

7. The method for determining the return period of historical floods based on multi-section flood correlation according to claim 1, characterized in that, Step 4 specifically involves: Step 4-1: Construct the annual maximum flood peak flow sequence for the benchmark section. The annual maximum flood peak flow sequence includes historical flood survey data and measured data. Step 4-2: Use P-III type curves to perform frequency analysis and linear fitting on the annual maximum flood peak flow sequence; Step 4-3: When the goodness of fit meets the requirements, the precise statistical return period corresponding to the first major historical flood can be read from the fitted frequency curve. .

8. The method for determining the return period of historical floods based on multi-section flood correlation according to claim 7, characterized in that, During the linear fitting process, the coefficient of variation is adjusted while maintaining reasonable key parameters. and skewness coefficient ,Keep =3.

00.

9. The method for determining the return period of historical floods based on multi-section flood correlation according to claim 8, characterized in that, The principle of alignment is to control the relative deviation between the data from the first major historical flood and the theoretical frequency curve within a set range.

10. The method for determining the return period of historical floods based on multi-section flood correlation according to claim 9, characterized in that, The relative deviation is no greater than 5%; In step 4-3, the goodness of fit R 2 The requirement is not less than 0.95.

Images