A method and system for screening a typical hydrological station in a watershed considering space-time characteristics

Abstract

The application discloses a kind of considering space-time characteristics method and system of screening basin typical hydrological station, comprising: obtaining basic data, and the original flow sequence of all candidate hydrological stations is basic data;According to basic data, based on the four-order construction mechanism of area weighting, underlying surface correction, double-factor verification and adaptive feedback regulation, the flow sequence of basin datum is constructed;The weighted integrated result of flow process of each candidate hydrological station in the basin range is basin datum flow sequence, which reflects the overall hydrological characteristics of basin;Construct multi-dimensional evaluation index system;Evaluation index system includes time consistency index and spatial representation ability index;According to the flow sequence of basin datum and evaluation index system, the weighted grey correlation degree of each candidate hydrological station is calculated based on grey correlation degree evaluation model;The weighted grey correlation degree of all candidate hydrological stations is sorted from large to small, and the candidate hydrological station with the maximum weighted grey correlation degree is selected as the basin typical hydrological station.The application realizes the accurate screening of typical station.

Description

Technical Field

[0001] This invention relates to the field of watershed hydrological station selection technology, specifically to a method and system for selecting typical watershed hydrological stations that considers spatiotemporal characteristics. Background Technology

[0002] In fields such as hydrological analysis, water resources planning, and watershed hydropower scheduling, accurately grasping the hydrological situation of a watershed is the foundation for scientific decision-making. This typically relies on long-term observation data from numerous hydrological stations deployed within the watershed. However, directly using data from all stations in practice faces significant challenges: on the one hand, the large number of stations leads to high costs and low efficiency in data collection, processing, and analysis, and information redundancy can easily obscure core patterns; on the other hand, different regions within the watershed have varying geographical, climatic, and underlying surface conditions, resulting in different hydrological processes recorded by each station (such as runoff timing, abundance and scarcity characteristics, and fluctuation patterns). Arbitrarily selecting a few typical hydrological stations may lack scientific basis and fail to represent the overall hydrological characteristics of the watershed, thus leading to biases or one-sidedness in scheduling strategies or hydrological conclusions based on data from that station.

[0003] Existing technologies for selecting watershed hydrological stations generally suffer from two main drawbacks: First, relying on manual experience to select mainstream or outlet stations lacks quantitative basis, easily leading to the selected stations failing to characterize the overall watershed hydrological heterogeneity. Second, existing quantitative methods mostly employ single-dimensional indicators (such as area ratio and linear correlation coefficient), failing to consider the nonlinear morphological characteristics of flow processes and unable to quantify the spatiotemporal differences between wet and dry seasons. This is especially true for non-stationary runoff sequences, making it difficult to distinguish the essential difference between "consistent trends but amplitude deviations" and "similar amplitudes but opposite trends." Furthermore, traditional clustering methods have stringent requirements for data length, making them unsuitable for short-sequence station scenarios.

[0004] In view of the above, this application is hereby submitted. Summary of the Invention

[0005] The technical problems to be solved by this invention are: firstly, existing methods rely on manual experience for selection, resulting in the selection of stations that cannot represent the hydrological heterogeneity of the entire watershed; secondly, most methods rely on a single indicator, leading to insufficient representativeness and an inability to distinguish between "opposite trends" and "no trend correlation," thus failing to achieve accurate selection of typical stations. This invention provides a method and system for selecting typical hydrological stations in a watershed that considers spatiotemporal characteristics. It constructs a baseline flow sequence for the watershed through a two-factor validation and adaptive feedback adjustment mechanism, and combines a quantitative indicator system considering spatiotemporal characteristics with an evaluation model based on grey relational analysis to achieve accurate selection of typical stations.

[0006] This invention is achieved through the following technical solution:

[0007] In a first aspect, the present invention provides a method for selecting typical hydrological stations in a watershed that considers spatiotemporal characteristics, the method comprising:

[0008] Obtain basic data, which consists of the original flow sequences of all candidate hydrological stations;

[0009] Based on the basic data, a basin baseline flow sequence is constructed using a four-order construction mechanism of area weighting, underlying surface correction, two-factor verification, and adaptive feedback adjustment. The basin baseline flow sequence is a weighted integration result of the flow processes of each candidate hydrological station within the basin, which reflects the overall hydrological characteristics of the basin.

[0010] Construct a multi-dimensional evaluation index system; the evaluation index system includes time consistency indicators and spatial representation ability indicators;

[0011] Based on the basin baseline flow sequence and evaluation index system, the weighted grey relational degree of each candidate hydrological station is calculated using a grey relational evaluation model.

[0012] All candidate hydrological stations are sorted by weighted grey relational degree from largest to smallest, and the candidate hydrological station with the largest weighted grey relational degree is selected as the typical hydrological station of the watershed.

[0013] Furthermore, the method also includes:

[0014] The basic data is preprocessed to obtain preprocessed basic data; the preprocessing includes missing value imputation, outlier removal and sequence consistency test.

[0015] Furthermore, based on a fourth-order construction mechanism of area weighting, underlying surface correction, two-factor verification, and adaptive feedback adjustment, a watershed baseline discharge sequence is constructed, including:

[0016] By controlling the area ratio and the underlying surface correction coefficient, the flow of all candidate stations is weighted and integrated to construct the watershed baseline flow sequence;

[0017] The reference validity of the basin baseline flow sequence is verified based on the two-factor validation mechanism to obtain a validated basin baseline flow sequence; the two-factor validation mechanism includes physical consistency verification and external data fit verification.

[0018] Furthermore, based on a fourth-order construction mechanism of area weighting, underlying surface correction, two-factor verification, and adaptive feedback adjustment, a watershed baseline discharge sequence is constructed, which also includes:

[0019] If any of the two-factor validation mechanisms is not satisfied, error diagnosis and dynamic weight correction are performed based on the benchmark adaptive feedback adjustment mechanism.

[0020] The benchmark adaptive feedback adjustment mechanism includes:

[0021] Step 1: Let the number of iterations be... The initial baseline flow sequence was calculated. ;

[0022] Step 2: Based on the initial baseline flow sequence Calculate the correlation coefficient and trend deviation ,when If the condition is met, proceed to step 7; otherwise, proceed to step 3.

[0023] Step 3: Determine the iteration number Is it less than the preset threshold? If yes, proceed to step 4; otherwise, remove the site with the greatest negative impact on the verification metrics (i.e., the site that would increase the relevance coefficient the most after removal), and recalculate the weights. ,make Proceed to step 6;

[0024] Step 4: Calculate the average residual for each station If all sites If the signs are the same, it is determined to be a global deviation; otherwise, it is determined to be a local deviation; the average residual... The calculation method is as follows:

[0025] ;

[0026] ;

[0027] in, The total number of all candidate hydrological stations in the basin; For the first The flow sequences of the candidate hydrological stations are in Values ​​for the time period; For the first The candidate hydrological stations at time The residual;

[0028] Step 5: When the deviation is a global deviation, a proportional adjustment coefficient is introduced. The initial baseline traffic sequence is scaled as a whole; when the deviation is a local deviation, an entropy correction factor is introduced to reduce the weight of stations in that area in the baseline construction. ;

[0029] Step 6: Based on the new weights Or proportional adjustment coefficient A new baseline flow sequence is obtained through calculation. ,make Proceed to step 2;

[0030] Step 7: Let Output the basin baseline flow sequence .

[0031] Furthermore, the expression for the basin baseline discharge sequence is:

[0032] ;

[0033] ;

[0034] In the formula, A baseline flow sequence for the watershed, reflecting the overall characteristics of the watershed; For the first The flow sequences of the candidate hydrological stations are in Values ​​for the time period; For the first The area weighting coefficients for each candidate hydrological station are determined based on the proportion of the catchment area controlled by each candidate hydrological station to the total catchment area, and a correction factor for the underlying surface is introduced. The underlying surface correction coefficient is obtained through remote sensing interpretation and is calculated by weighting based on vegetation coverage, soil type and slope distribution within the control basin of the candidate hydrological stations. Candidate hydrological stations The controlled catchment area; Index the candidate hydrological stations; This represents the total number of candidate hydrological stations in the basin.

[0035] Furthermore, physical consistency verification is performed by calculating the correlation coefficient between the baseline runoff depth and the watershed surface rainfall depth; the expression for the correlation coefficient is:

[0036] ;

[0037] ;

[0038] ;

[0039] ;

[0040] ;

[0041] In the formula, The correlation coefficient between baseline runoff depth and watershed surface rainfall depth; , They are respectively Baseline runoff depth and watershed surface rainfall depth for the specified time period; This serves as the baseline sequence for the overall flow characteristics of the watershed, i.e., the baseline flow sequence of the watershed. Drainage area; The time period is long; For the first The Thiessen polygon weighting coefficient of each rain gauge station is determined by the proportion of the control area corresponding to that station to the total watershed area. For the first Each rain gauge at time The observed rainfall; Index of the numbering of the watershed rainfall monitoring stations; This represents the total number of rainfall monitoring stations in the watershed. , These are the average values ​​of the baseline runoff depth and the watershed surface rainfall depth, respectively.

[0042] Furthermore, external data consistency verification is performed by calculating the trend deviation rate between the baseline runoff accumulation and the land water storage change series; the expression for the trend deviation rate is:

[0043] ;

[0044] ;

[0045] In the formula, The trend deviation rate of the baseline runoff accumulation and the change in terrestrial water storage TWSA series; Baseline runoff cumulative volume; for Baseline flow rate for the time period The time period is long; It represents the long-term trend of a time series and is obtained through linear regression.

[0046] Furthermore, the temporal consistency index is used to assess the degree of agreement between the original flow sequence of candidate hydrological stations and the baseline flow sequence of the watershed over time; the temporal consistency index includes morphological trend similarity index and trend correlation index.

[0047] Spatial characterization indicators are used to assess the coverage representativeness of hydrological characteristic values ​​of candidate hydrological stations in the spatial distribution of the watershed; spatial characterization indicators include water level representativeness indicators and the correlation between wet and dry seasons.

[0048] Furthermore, based on the basin baseline flow sequence and evaluation index system, the weighted grey relational degree of each candidate hydrological station is calculated using a grey relational evaluation model, including:

[0049] The basin baseline discharge series is used as an ideal reference series;

[0050] Based on the evaluation index system, calculate the evaluation index matrix of the flow sequence of each candidate station and the baseline flow sequence of the watershed;

[0051] The range normalization method is used to normalize the original index values ​​in the evaluation index matrix to obtain the normalized evaluation index matrix.

[0052] For each candidate hydrological station and each evaluation index, calculate its grey relational coefficient with the ideal reference sequence;

[0053] The weights of each evaluation index are determined using the entropy weight method. Based on the weights and the grey relational coefficient, the weighted grey relational degree of each candidate hydrological station is calculated.

[0054] Furthermore, the weighted grey relational scores of all candidate hydrological stations are sorted from largest to smallest, and the candidate hydrological station with the largest weighted grey relational score is selected as the typical hydrological station of the basin, including:

[0055] Sort all candidate hydrological stations by weighted grey relational degree from largest to smallest;

[0056] If a typical hydrological station for the basin is required, the candidate hydrological station with the highest weighted grey relational degree is selected as the typical hydrological station for the basin.

[0057] If multiple typical hydrological stations in the river basin are required, select the top stations in descending order of weighted grey relational analysis. A candidate hydrological station, or a set of thresholds to determine the typical set of hydrological stations.

[0058] Secondly, the present invention provides a system for screening typical hydrological stations in a watershed that considers spatiotemporal characteristics, the system comprising:

[0059] The acquisition unit is used to acquire basic data, which consists of the original flow sequences of all candidate hydrological stations.

[0060] The watershed baseline discharge sequence construction unit is used to construct the watershed baseline discharge sequence based on basic data and a fourth-order construction mechanism of area weighting, underlying surface correction, two-factor verification and adaptive feedback adjustment. The watershed baseline discharge sequence is the weighted integration result of the discharge process of each candidate hydrological station within the watershed.

[0061] The evaluation index construction unit is used to construct a multi-dimensional evaluation index system; the evaluation index system includes time consistency indicators and spatial representation ability indicators.

[0062] The grey relational evaluation unit is used to calculate the weighted grey relational degree of each candidate hydrological station based on the basin baseline flow sequence and evaluation index system and the grey relational evaluation model.

[0063] The filtering unit is used to sort all candidate hydrological stations by weighted grey relational degree from largest to smallest, and select the candidate hydrological station with the largest weighted grey relational degree as the typical hydrological station of the watershed.

[0064] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0065] This invention provides a method and system for screening typical hydrological stations in a watershed that considers spatiotemporal characteristics. By constructing a watershed baseline flow sequence verified by a two-factor validation mechanism, and combining a quantitative index system that considers spatiotemporal characteristics with an evaluation model based on grey relational analysis, the method achieves accurate screening of typical stations.

[0066] 1. Breakthrough in the objectivity of benchmark construction: For the first time, a dual-factor verification mechanism of "physical consistency + external data" and adaptive feedback adjustment logic are introduced, which solves the problem of "emphasizing mathematical integration and neglecting physical verification" in traditional benchmark sequences, and ensures that the benchmark sequence truly reflects the water cycle law of the basin.

[0067] 2. Comprehensive breakthrough in evaluation dimensions: The constructed indicator system covers four aspects simultaneously: process form, rate of change, water volume level, and encounter between abundant and dry seasons, making up for the technical blind spot of existing methods that cannot distinguish between "opposite trends" and "no trend correlation".

[0068] 3. Breakthrough in robustness of evaluation results: The entropy weight method-grey relational degree combination model avoids the arbitrariness of subjective weighting and has significantly better adaptability to short and non-stationary sequences than traditional clustering methods, which can support the screening needs of typical stations in watersheds of different sizes.

[0069] 4. It realizes the connection between scientific research benchmarks and engineering applications: This application is the first to clearly distinguish the different functional positions of "basin benchmark flow sequence" and "typical hydrological station": the former serves as a physically verifiable whole basin reference scale, while the latter serves as a monitoring node that can be deployed in an engineering manner, which solves the problem of existing methods having "evaluation standards but no implementation carrier". Attached Figure Description

[0070] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0071] Figure 1 This is a flowchart of a method for selecting typical hydrological stations in a watershed that considers spatiotemporal characteristics, according to the present invention.

[0072] Figure 2 This is a detailed flowchart of a method for selecting typical hydrological stations in a watershed that considers spatiotemporal characteristics, according to the present invention.

[0073] Figure 3 This diagram illustrates the adaptive feedback adjustment mechanism of the overall baseline flow sequence in a watershed, which is part of the present invention's method for selecting typical hydrological stations in a watershed that considers spatiotemporal characteristics.

[0074] Figure 4 This is a flowchart of the typical hydrological station evaluation process based on grey relational analysis, which is a method for selecting typical hydrological stations in a watershed that considers spatiotemporal characteristics according to the present invention.

[0075] Figure 5 This is a structural block diagram of a typical hydrological station screening system for a watershed that takes into account spatiotemporal characteristics, according to the present invention. Detailed Implementation

[0076] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0077] The core idea of ​​this invention is to evaluate the closeness of the flow process of candidate stations to the overall flow and hydrological characteristics of the basin from both temporal and spatial dimensions based on the constructed basin baseline flow sequence, and to calculate the weighted grey relational degree of each candidate hydrological station based on the grey relational evaluation model in order to screen typical hydrological stations in the basin.

[0078] Example 1

[0079] like Figure 1 and Figure 2 As shown, this invention provides a method for selecting typical hydrological stations in a watershed that considers spatiotemporal characteristics. The method includes:

[0080] S1. Obtain basic data, which consists of the original flow sequences of all candidate hydrological stations; preprocess the basic data to obtain preprocessed basic data.

[0081] Specifically, preprocessing includes missing value imputation, outlier removal, and sequence consistency verification to ensure the reliability and continuity of the original traffic sequence.

[0082] S2, based on the basic data, a watershed baseline flow sequence is constructed using a four-order construction mechanism of area weighting, underlying surface correction, two-factor verification, and adaptive feedback adjustment.

[0083] Before evaluating the representativeness of candidate hydrological stations, it is necessary to first clarify the comparison benchmark. This invention proposes to construct a basin benchmark flow sequence that reflects the overall hydrological characteristics of the basin. This sequence is defined as a weighted integration of the flow processes of each candidate hydrological station within the basin, serving as a unified reference standard for representativeness assessment. Unlike existing technologies that directly use the outflow station flow as the representative value of the basin, the basin benchmark flow sequence constructed in this application is not the measured value of a single candidate hydrological station, but rather a four-order construction mechanism based on area weighting, underlying surface correction, dual-factor verification, and adaptive feedback adjustment. First, the flow of all candidate hydrological stations is weighted and integrated by controlling the area proportion and underlying surface correction coefficient. Second, the rationality of the benchmark is verified through a dual verification of "precipitation-runoff physical consistency verification + GRACE water storage external verification." If the verification fails, an adaptive feedback adjustment mechanism is triggered until the basin benchmark flow sequence simultaneously satisfies both physical laws and external observation constraints. This basin benchmark flow sequence retains the physical mechanism of basin runoff generation and elimination of interference from single-station observation errors and local extreme events, providing an objective reference benchmark for subsequent representativeness assessment.

[0084] Specifically, step S2 includes:

[0085] S21, by weighting and fusing the flows of all candidate stations using the control area ratio and underlying surface correction coefficient, a watershed baseline flow sequence is constructed; the expression for the watershed baseline flow sequence is:

[0086] ; (1)

[0087] ; (2)

[0088] In the formula, A baseline flow sequence for the watershed, reflecting the overall characteristics of the watershed; For the first The flow sequences of the candidate hydrological stations are in Values ​​for the time period; For the first The area weighting coefficients for each candidate hydrological station are determined based on the proportion of the catchment area controlled by each candidate hydrological station to the total catchment area, and a correction factor for the underlying surface is introduced. The underlying surface correction coefficient is obtained through remote sensing interpretation and is calculated by weighting based on vegetation coverage, soil type and slope distribution within the control basin of the candidate hydrological stations. Candidate hydrological stations The controlled catchment area; Index the candidate hydrological stations; This represents the total number of candidate hydrological stations in the basin.

[0089] S22, the reference validity of the basin baseline flow sequence is verified based on the two-factor validation mechanism to obtain a validated basin baseline flow sequence; the two-factor validation mechanism includes physical consistency verification and external data fit verification.

[0090] To ensure This invention introduces a two-factor validation mechanism to verify the reference value of the basin baseline flow sequence, which can truly reflect the overall flow situation of the basin rather than just a mathematical accumulation of data.

[0091] (1) Physical consistency verification

[0092] Test the basin baseline flow sequence The relationship between the baseline runoff depth and the watershed rainfall input is analyzed. The correlation coefficient between the baseline runoff depth and the watershed surface rainfall depth is calculated. If this correlation coefficient is greater than a preset threshold of 0.85 (which can be adjusted according to watershed characteristics), it proves that the watershed baseline flow sequence conforms to the physical laws of the water cycle and has physical reference value. The calculation methods for baseline runoff depth, watershed surface rainfall depth, and correlation coefficient are as follows:

[0093] ; (3)

[0094] ; (4)

[0095] ; (5)

[0096] ; (6)

[0097] ; (7)

[0098] In the formula, The correlation coefficient between baseline runoff depth and watershed surface rainfall depth; , They are respectively Baseline runoff depth and watershed surface rainfall depth for the specified time period; This serves as the baseline sequence for the overall flow characteristics of the watershed, i.e., the baseline flow sequence of the watershed. Drainage area; The time period is long; For the first The Thiessen polygon weighting coefficient of each rain gauge station is determined by the proportion of the control area corresponding to that station to the total watershed area. For the first Each rain gauge at time The observed rainfall; Index of the numbering of the watershed rainfall monitoring stations; This represents the total number of rainfall monitoring stations in the watershed. , These are the average values ​​of the baseline runoff depth and the watershed surface rainfall depth, respectively.

[0099] (2) External data consistency verification

[0100] Cross-validation of the baseline was performed using remote sensing or reanalysis data independent of site observations. The trend of terrestrial water storage change (TWSA) was retrieved from GRACE satellite gravity data. If the trend deviation between the two is within ±5%, it demonstrates that the baseline flow series for the watershed can represent the true water storage change at the watershed scale and has objective reference value.

[0101] External data consistency verification is achieved by calculating baseline runoff accumulation. The trend deviation rate is used to verify this against the trend deviation rate of the terrestrial water storage change series (TWSA series); the expression for the trend deviation rate is:

[0102] ; (8)

[0103] ; (9)

[0104] In the formula, The trend deviation rate of the baseline runoff accumulation and the change in terrestrial water storage TWSA series; Baseline runoff cumulative volume; for Baseline flow rate for the time period The time period is long; It represents the long-term trend of a time series and is obtained through linear regression.

[0105] S23, If any of the two-factor validation mechanisms is not satisfied, error diagnosis and dynamic weight correction are performed based on the benchmark adaptive feedback adjustment mechanism.

[0106] In practical engineering, due to potential systematic errors in historical observation data, uneven distribution of stations, or interference from human activities, the initially constructed watershed baseline flow sequence may not pass the aforementioned physical consistency verification or external conformity verification on the first attempt. To ensure the authority of the baseline, when the watershed baseline flow sequence... When the correlation coefficient with watershed rainfall depth is below the threshold of 0.85, or the deviation from the trend of water storage change on the GRACE satellite exceeds ±5%, this invention designs a benchmark adaptive feedback adjustment mechanism, the process of which is as follows: Figure 3 As shown, the specific steps are as follows:

[0107] Step 1: Let the number of iterations be... The initial baseline flow sequence was calculated. ;

[0108] Step 2: Based on the initial baseline flow sequence The correlation coefficient is calculated based on equations (3)-(9). and trend deviation ,when If the condition is met, proceed to step 7; otherwise, proceed to step 3.

[0109] Step 3: Determine the iteration number Is it less than the preset threshold? If yes, proceed to step 4; otherwise, remove the site with the greatest negative impact on the verification metrics (i.e., the site that would increase the relevance coefficient the most after removal), and recalculate the weights. ,make Proceed to step 6;

[0110] Step 4: Calculate the average residual for each station If all sites If the signs are the same, it is determined to be a global bias; otherwise, it is determined to be a local bias; mean residual The calculation method is as follows:

[0111] ; (10)

[0112] ; (11)

[0113] in, The total number of all candidate hydrological stations in the basin; For the first The flow sequences of the candidate hydrological stations are in Values ​​for the time period; For the first The candidate hydrological stations at time The residual;

[0114] Step 5: When the deviation is a global deviation, a proportional adjustment coefficient is introduced. The initial baseline traffic sequence is scaled as a whole; when the deviation is a local deviation, an entropy correction factor is introduced to reduce the weight of stations in that area in the baseline construction. ;

[0115] Step 6: Based on the new weights Or proportional adjustment coefficient A new baseline flow sequence is obtained through calculation. ,make Proceed to step 2;

[0116] Step 7: Let Output the basin baseline flow sequence .

[0117] S3, Construct a multi-dimensional evaluation index system; the evaluation index system includes time consistency indicators and spatial representation ability indicators;

[0118] Based on two core dimensions—temporal consistency and spatial representativeness—four quantitative indicators are constructed to comprehensively evaluate the degree of matching between candidate hydrological stations and the overall basin flow characteristic benchmark. The temporal consistency indicator assesses the degree of agreement between the original flow sequences of candidate hydrological stations and the basin benchmark flow sequences over time, including morphological trend similarity and correlation of change trends. The spatial representation capability indicator assesses the coverage representativeness of the hydrological characteristic values ​​of candidate hydrological stations in the spatial distribution of the basin, including the representativeness of water volume levels and the correlation between wet and dry seasons, reflecting the station's ability to characterize the hydrological heterogeneity of different regions within the basin.

[0119] S31, Time Consistency Index

[0120] (1) Pattern trend similarity index

[0121] To address the limitation of traditional Euclidean distance, which only reflects numerical deviations and fails to capture the degree of morphological alignment, this application employs a sliding cross-correlation-Z-score normalization combination algorithm: first, the sequences are normalized to eliminate magnitude differences; then, the maximum cross-correlation coefficient is calculated through a sliding window, ultimately mapping to morphological similarity distance. This index possesses scaling invariance and displacement invariance, effectively identifying sequences with "consistent peak shapes but phase shifts," avoiding misjudgments caused by the lag in flood propagation. The calculation process is as follows:

[0122] ;(12)

[0123] ; (13)

[0124] ;(14)

[0125] ; (15)

[0126] ; (16)

[0127] In the formula, and For two lengths Time series; and These are the mean and standard deviation of the sequence, respectively. Represents a sequence Relative to sequence The sliding step size (positive and negative indicate the sliding direction); Cross-correlation sequences The position index; It is the sum of the dot products of the overlapping parts after sliding; and Each has zero-hysteresis autocorrelation; The cross-correlation sequence after coefficient normalization has a value range of [-1, 1], and its maximum value reflects the maximum similarity of the morphological trends of the two sequences. The range is [0,2], where 0 indicates that the shapes are completely similar and 2 indicates that they are completely opposite.

[0128] (2) Correlation indicators of changing trends

[0129] Besides the static water volume, the rate of change in flow rate is also a crucial parameter for hydrological station scheduling. Considering that traditional methods based on absolute numerical differences are highly sensitive to outliers and cannot accurately distinguish between the fundamentally different situations of "completely opposite trends" and "completely unrelated trends," this invention further employs the Spearman rank correlation coefficient, a statistical measure specifically used to assess the consistency of monotonic trends, as the core metric. This method does not rely on the absolute magnitude of the original values; it judges the degree of trend matching solely through the relative rank of the data. It naturally possesses resistance to outlier interference, has clear statistical significance, and is more robust. Its calculation formula is as follows:

[0130] ; (17)

[0131] ; (18)

[0132] In the formula, Spearman's rank correlation coefficient is a core statistic for measuring the consistency of the monotonic trends of two sequences. Its value ranges from [-1, 1]. When the trend of the baseline flow series is completely consistent with that of the flow series to be evaluated, When the trend of change of the two sequences is completely opposite, This indicates that the changing trends of the two sequences are uncorrelated; The total length of the slope sequence is equal to the length of the original flow sequence minus the time step, i.e. ,in This represents the total number of time points in the original flow sequence. The time step for calculating the rate of change of flow (e.g., 1 hour, 1 day). This is a time index, with a value range of [value range missing]. , corresponding to the time of calculation for each slope; For the first The difference in rank at time, The slope of the baseline flow rate The rank (i.e., the position number after sorting by numerical value) in the entire baseline slope sequence. The slope of the flow rate to be evaluated The rank within the entire slope sequence to be evaluated; This is a fixed constant coefficient in the Spearman rank correlation coefficient formula, derived from the mathematical properties of the rank difference, and does not need to be adjusted according to actual data.

[0133] S32, Spatial Representation Ability Index

[0134] (1) Representative indicators of water quantity level

[0135] A good, typical hydrological station should have a multi-year average water volume at the "mainstream" level of the watershed, rather than an extreme value. Traditional simple difference methods measure similarity only through absolute water volume differences, neglecting the magnitude of the baseline water volume itself. This leads to a lack of comparability between watersheds of different sizes and fails to intuitively reflect the degree of similarity. Therefore, this invention employs a normalization method based on relative differences. It measures the representativeness of water volume levels by calculating the relative deviation of the total water volume between two sequences. The calculation formula is as follows:

[0136] ; (19)

[0137] In the formula, Representative indicators of water volume levels; Represents the baseline flow series of the watershed Traffic volume during specific time periods; Indicates the test hydrological station Traffic volume during specific time periods; Indicates the number of time periods; The smaller the value, the more similar the water volume of the two.

[0138] (2) Correlation index between abundance and scarcity

[0139] In hydrological scheduling, simply knowing the correlation between flow rates is insufficient; more importantly, it's crucial to ensure synchronization between dry and wet seasons. To this end, this invention introduces set pair analysis (SPA) theory. Unlike the traditional Pearson correlation coefficient, the correlation degree can distinguish the attribute differences behind "simultaneous increases and decreases." First, the Weibull formula is used to calculate empirical frequencies to divide the three levels of wet, normal, and dry seasons.

[0140] ; (20)

[0141] Then, a number of connections was built. :

[0142] ; (twenty one)

[0143] In the formula, It is the total number of elements in the runoff sequence. This represents the number of elements with the same sign in two sequences, therefore Indicates the same degree; This indicates the number of elements whose runoff abundance / shortness level difference is 1. The identifier representing difference 1, This indicates the proportion of difference 1 in the total population and its related characteristics; This indicates the number of elements whose runoff abundance / shortness level differs by 2. For identifiers with a difference level of 2, This indicates the proportion and related characteristics of elements with a runoff abundance / dampness level of 2 in the total population; Indicates the difference in runoff abundance / scarcity levels The number of elements at level 3 (opposite). These are identifiers that represent opposition. This indicates the proportion and related characteristics of opposing elements. In practical applications, based on experience, it is generally believed that... , , It is reasonable; The larger the value, the higher the correlation between the abundance and scarcity of the two sequences, and the stronger the representativeness; conversely, the lower the value, the weaker the representativeness.

[0144] S4. Based on the basin baseline flow sequence and evaluation index system, calculate the weighted grey relational degree of each candidate hydrological station using the grey relational degree evaluation model.

[0145] In step S3 above, the four indicators characterize the degree of similarity between the candidate hydrological stations and the overall hydrological characteristics of the watershed from different physical perspectives. Simply averaging or subjectively weighting the four indicators fails to reflect the differentiated contributions of each indicator in the comprehensive evaluation and makes it difficult to guarantee the objectivity and robustness of the screening results. Therefore, this invention further introduces a Grey Relational Analysis (GRA) model to quantitatively calculate the degree of similarity between the flow sequence of each candidate hydrological station and the baseline flow sequence of the watershed, thereby selecting the optimal typical hydrological station. The specific process is as follows: Figure 4 As shown, the specific steps are as follows:

[0146] S41, Construct a virtual ideal hydrological station. To facilitate the calculation of grey relational degree, the watershed baseline discharge sequence verified by a two-factor validation mechanism is used as the ideal reference sequence;

[0147] S42, Calculate the evaluation index values. Based on the evaluation index system, calculate the evaluation index matrix of the flow sequence at each candidate station and the baseline flow sequence of the watershed. ;

[0148] ; (twenty two)

[0149] In the formula, , , , Sites The indicators include morphological trend similarity, trend correlation, water volume level representativeness, and the degree of correlation between abundant and dry seasons.

[0150] S43, Dimensionless Data Processing. To avoid the influence of dimensions and numerical magnitude on the evaluation results, the range normalization method is used to normalize the original index values ​​in the evaluation index matrix, resulting in a normalized evaluation index matrix.

[0151] For the The first hydrological station The normalized values ​​of the following indicators are:

[0152] Positive indicators:

[0153] ; (twenty three)

[0154] Negative indicators:

[0155] ; (twenty four)

[0156] in, Positive / negative indicators For the first The first candidate hydrological station One evaluation indicator, For all candidate hydrological stations The minimum value of each evaluation indicator. For all candidate hydrological stations The maximum value of each evaluation indicator.

[0157] The normalized evaluation index matrix is ​​as follows:

[0158] ; (25)

[0159] After the above transformation, the evaluation index of the ideal reference sequence becomes The index values ​​of all candidate hydrological stations Furthermore, a larger value indicates better performance.

[0160] S44, calculate the grey relational coefficient. For each candidate hydrological station. and each evaluation indicator Calculate its grey relational coefficient with the ideal reference sequence. ;

[0161] ; (26)

[0162] In the formula, The resolution coefficient is typically set to 0.5, and its function is to adjust for differences between correlation coefficients. It is the global minimum absolute difference (usually 0). It represents the global maximum absolute difference.

[0163] S45, Calculate the weighted grey relational degree. The entropy weight method is used to determine the weights of each evaluation index. Let the weight vector of the four indicators be... ,satisfy Based on the weights and grey relational coefficients, the weighted grey relational degree of each candidate hydrological station is calculated.

[0164] No. Weighted grey relational degree of each candidate hydrological station for:

[0165] ;(27)

[0166] In the formula, The closer the value is to 1, the closer the flow process of the hydrological station is to the overall characteristics of the watershed, and the stronger its representativeness.

[0167] S5. Sort all candidate hydrological stations by weighted grey relational degree from largest to smallest, and select the candidate hydrological station with larger weighted grey relational degree as the typical hydrological station of the watershed.

[0168] Specifically, step S5 includes:

[0169] Weighted grey relational analysis of all candidate hydrological stations Sort by size from largest to smallest; if only one typical hydrological station is needed for the basin, select the candidate station with the highest weighted grey relational degree as the typical hydrological station for the basin; if multiple typical hydrological stations are actually needed (e.g., reflecting the characteristics of different sub-regions), select the stations in descending order of weighted grey relational degree. A candidate hydrological station, or a threshold (such as...) Determine the set of typical hydrological stations.

[0170] The basin baseline flow sequence constructed in step S2 of this invention is a full-element reference sequence obtained by weighted fusion of observation data from all candidate hydrological stations in the entire basin, mainly used as a "benchmark" for representativeness assessment; while the typical hydrological stations in the basin selected in step S5 are used to achieve low-cost and high-efficiency hydrological monitoring and scheduling decisions in subsequent actual operations. If the basin baseline sequence is directly used for daily hydrological analysis, reservoir scheduling, or drought and flood early warning, the following problems will be faced:

[0171] (1) Excessive data dependence: It is necessary to obtain real-time data from all stations in the entire basin simultaneously. The computational difficulty of data transmission, cleaning and fusion is high, making it difficult to meet the timeliness requirements of flood control scheduling, etc.

[0172] (2) Difficulty in project implementation: In actual watershed management, the cost of maintaining high-density monitoring of all stations in the long term is extremely high, and it is easily affected by station shutdowns and data loss;

[0173] (3) Redundant information interference: The whole basin sequence contains a large amount of local heterogeneous information, which is not conducive to grasping the dominant hydrological patterns.

[0174] Therefore, by selecting typical stations, this invention simplifies "multi-station fusion across the entire basin" to "single-station or few-station application" while ensuring high consistency with the benchmark sequence, achieving a balance between accuracy and cost, and possessing significant engineering practical value.

[0175] Example 2

[0176] like Figure 5 As shown, the difference between this embodiment and Embodiment 1 is that this embodiment provides a system for selecting typical hydrological stations in a watershed that considers spatiotemporal characteristics. This system corresponds one-to-one with the method for selecting typical hydrological stations in a watershed that considers spatiotemporal characteristics in Embodiment 1. The system includes:

[0177] The acquisition unit is used to acquire basic data, which consists of the original flow sequences of all candidate hydrological stations.

[0178] The watershed baseline discharge sequence construction unit is used to construct the watershed baseline discharge sequence based on basic data and a fourth-order construction mechanism of area weighting, underlying surface correction, two-factor verification and adaptive feedback adjustment. The watershed baseline discharge sequence is the weighted integration result of the discharge process of each candidate hydrological station within the watershed.

[0179] The evaluation index construction unit is used to construct a multi-dimensional evaluation index system; the evaluation index system includes time consistency indicators and spatial representation ability indicators.

[0180] The grey relational evaluation unit is used to calculate the weighted grey relational degree of each candidate hydrological station based on the basin baseline flow sequence and evaluation index system and the grey relational evaluation model.

[0181] The filtering unit is used to sort all candidate hydrological stations by weighted grey relational degree from largest to smallest, and select the candidate hydrological stations with larger weighted grey relational degrees as typical hydrological stations in the basin.

[0182] The execution process of each unit can be carried out according to the steps of the typical hydrological station selection method for watershed considering spatiotemporal characteristics in Example 1, and will not be described in detail in this example.

[0183] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0184] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0185] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0186] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0187] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for selecting typical hydrological stations in a watershed considering spatiotemporal characteristics, characterized in that, The method includes: Acquire basic data, which consists of the original flow sequences of all candidate hydrological stations; Based on the aforementioned basic data, a basin baseline flow sequence is constructed using a fourth-order construction mechanism that incorporates area weighting, underlying surface correction, two-factor verification, and adaptive feedback adjustment. The basin baseline flow sequence is a weighted integration result of the flow processes at each candidate hydrological station within the basin. Construct a multi-dimensional evaluation index system; the evaluation index system includes time consistency indicators and spatial representation ability indicators; Based on the basin baseline flow sequence and the evaluation index system, the weighted grey relational degree of each candidate hydrological station is calculated using the grey relational degree evaluation model. Sort all candidate hydrological stations by weighted grey relational degree from largest to smallest, and select the candidate hydrological station with the largest weighted grey relational degree as the typical hydrological station of the basin.

2. The method for selecting typical hydrological stations in a watershed considering spatiotemporal characteristics according to claim 1, characterized in that, A fourth-order construction mechanism based on area weighting, underlying surface correction, two-factor verification, and adaptive feedback adjustment is used to construct a watershed baseline discharge sequence, including: By controlling the area ratio and the underlying surface correction coefficient, the flow of all candidate stations is weighted and integrated to construct the watershed baseline flow sequence; The reference validity of the basin baseline flow sequence is verified based on a two-factor validation mechanism to obtain a validated basin baseline flow sequence; the two-factor validation mechanism includes physical consistency verification and external data matching verification.

3. The method for selecting typical hydrological stations in a watershed considering spatiotemporal characteristics according to claim 2, characterized in that, A fourth-order construction mechanism based on area weighting, underlying surface correction, two-factor verification, and adaptive feedback adjustment is used to construct the watershed baseline discharge sequence, which also includes: If any of the two-factor validation mechanisms is not satisfied, error diagnosis and dynamic weight correction are performed based on the benchmark adaptive feedback adjustment mechanism. The benchmark adaptive feedback adjustment mechanism includes: Step 1: Let the number of iterations be... The initial baseline flow sequence was calculated. ; Step 2: Based on the initial baseline flow sequence Calculate the correlation coefficient and trend deviation ,when If the condition is met, proceed to step 7; otherwise, proceed to step 3. Step 3: Determine the iteration number Is it less than the preset threshold? If yes, proceed to step 4; otherwise, remove the site with the greatest negative impact on the verification metrics (i.e., the site that would increase the relevance coefficient the most after removal), and recalculate the weights. ,make Proceed to step 6; Step 4: Calculate the average residual for each station If all sites If the signs are the same, it is determined to be a global deviation; otherwise, it is determined to be a local deviation; the average residual... The calculation method is as follows: ； ； in, The total number of all candidate hydrological stations in the basin; For the first The flow sequences of the candidate hydrological stations are in Values ​​for the time period; For the first The candidate hydrological stations at time The residual; Step 5: When the deviation is a global deviation, a proportional adjustment coefficient is introduced. The initial baseline traffic sequence is scaled as a whole; when the deviation is a local deviation, an entropy correction factor is introduced to reduce the weight of stations in that area in the baseline construction. ; Step 6: Based on the new weights Or proportional adjustment coefficient A new baseline flow sequence is obtained through calculation. ,make Proceed to step 2; Step 7: Let Output the basin baseline flow sequence .

4. The method for selecting typical hydrological stations in a watershed considering spatiotemporal characteristics according to claim 2, characterized in that, The expression for the basin baseline flow sequence is: ； ； In the formula, A baseline flow sequence for the watershed, reflecting the overall characteristics of the watershed; For the first The flow sequences of the candidate hydrological stations are in Values ​​for the time period; For the first The area weighting coefficients for each candidate hydrological station are determined based on the proportion of the catchment area controlled by each candidate hydrological station to the total catchment area, and a correction factor for the underlying surface is introduced. The underlying surface correction coefficient is obtained through remote sensing interpretation and is calculated by weighting based on vegetation coverage, soil type and slope distribution within the control basin of the candidate hydrological stations. Candidate hydrological stations The controlled catchment area; Index the candidate hydrological stations; This represents the total number of candidate hydrological stations in the basin.

5. The method for selecting typical hydrological stations in a watershed considering spatiotemporal characteristics according to claim 2, characterized in that, The physical consistency verification is performed by calculating the correlation coefficient between the baseline runoff depth and the watershed surface rainfall depth; the expression for the correlation coefficient is: ； ； ； ； ； In the formula, The correlation coefficient between baseline runoff depth and watershed surface rainfall depth; , They are respectively Baseline runoff depth and watershed surface rainfall depth for the specified time period; This serves as the baseline sequence for the overall flow characteristics of the watershed, i.e., the baseline flow sequence of the watershed. Drainage area; The time period is long; For the first The Thiessen polygon weighting coefficient of each rain gauge station is determined by the proportion of the control area corresponding to that station to the total watershed area. For the first Each rain gauge at time The observed rainfall; Index of the numbering of the watershed rainfall monitoring stations; This represents the total number of rainfall monitoring stations in the watershed. , These are the average values ​​of the baseline runoff depth and the watershed surface rainfall depth, respectively.

6. The method for selecting typical hydrological stations in a watershed considering spatiotemporal characteristics according to claim 2, characterized in that, The external data consistency verification is performed by calculating the trend deviation rate between the baseline runoff accumulation and the land water storage change series; the expression for the trend deviation rate is: ； ； In the formula, The trend deviation rate of the baseline runoff accumulation and the change in terrestrial water storage TWSA series; Baseline runoff cumulative volume; for Baseline flow rate for the time period The time period is long; It represents the long-term trend of a time series and is obtained through linear regression.

7. The method for selecting typical hydrological stations in a watershed considering spatiotemporal characteristics according to claim 1, characterized in that, The time consistency index includes morphological trend similarity index and change trend correlation index; The spatial characterization indicators include water level representativeness indicators and the correlation between abundant and scarce water conditions.

8. The method for selecting typical hydrological stations in a watershed considering spatiotemporal characteristics according to claim 1, characterized in that, Based on the basin baseline flow sequence and the evaluation index system, the weighted grey relational degree of each candidate hydrological station is calculated using a grey relational evaluation model, including: The basin baseline flow sequence is used as an ideal reference sequence; Based on the evaluation index system, calculate the evaluation index matrix of each candidate site flow sequence and the basin baseline flow sequence; The original index values ​​in the evaluation index matrix are normalized using the range normalization method to obtain the normalized evaluation index matrix. For each candidate hydrological station and each evaluation index, calculate its grey relational coefficient with the ideal reference sequence; The weights of each evaluation index are determined using the entropy weight method. Based on the weights and the grey relational coefficients, the weighted grey relational degree of each candidate hydrological station is calculated.

9. The method for selecting typical hydrological stations in a watershed considering spatiotemporal characteristics according to claim 1, characterized in that, All candidate hydrological stations are sorted by weighted grey relational degree from largest to smallest, and the candidate hydrological station with the largest weighted grey relational degree is selected as the typical hydrological station of the watershed, including: Sort all candidate hydrological stations by weighted grey relational degree from largest to smallest; If a typical hydrological station for the basin is required, then the candidate hydrological station with the largest weighted grey relational degree is selected as the typical hydrological station for the basin. If multiple typical hydrological stations in the watershed are required, then select the stations in descending order of the weighted grey relational degree. A candidate hydrological station, or a set of thresholds to determine the typical set of hydrological stations.

10. A system for selecting typical hydrological stations in a watershed that considers spatiotemporal characteristics, characterized in that, The system includes: The acquisition unit is used to acquire basic data, which is the original flow sequence of all candidate hydrological stations; The basin baseline flow sequence construction unit is used to construct the basin baseline flow sequence based on the aforementioned basic data and a fourth-order construction mechanism of area weighting, underlying surface correction, two-factor verification, and adaptive feedback adjustment; the basin baseline flow sequence is the weighted integration result of the flow processes of each candidate hydrological station within the basin. An evaluation index construction unit is used to construct a multi-dimensional evaluation index system; the evaluation index system includes time consistency indicators and spatial representation ability indicators. The grey relational evaluation unit is used to calculate the weighted grey relational degree of each candidate hydrological station based on the basin baseline flow sequence and the evaluation index system and the grey relational evaluation model. The filtering unit is used to sort all candidate hydrological stations by weighted grey relational degree from largest to smallest, and select the candidate hydrological station with the largest weighted grey relational degree as the typical hydrological station of the basin.

Images