Snow melting runoff calculation method based on meteorological element dominance analysis

Abstract

The invention discloses a meteorological element dominance analysis-based snow melting runoff calculation method. The method comprises the following specific steps of obtaining runoff data of a target hydrometric station; based on the runoff data of the target hydrometric station, constructing a target influence monthly factor matrix; constructing multiple models, and calculating to obtain the contribution rate of each target influence factor in different models to runoff; and calculating the snow melting runoff volume. According to the method, the partial correlation method is utilized to eliminate the influence of other influence factors on the runoff data of the target hydrometric station, and different influence factors are considered for the runoff volume of different months; a multi-model group of a monthly regression equation set sub-model and a plurality of target machine learning models is constructed, a weighted average multi-model integration contribution rate calculation mode is proposed by using fitting effects of different models to obtain the contribution rate of each target influence factor to runoff, and the non-universality of the contribution rate of the influence factor on a single model is effectively overcome; and the scientificity and the interpretability in the current snow melting runoff calculation method are enhanced.

Description

Technical Field

[0001] The present invention relates to the field of snowmelt runoff calculation. Specifically, it relates to a method for calculating snowmelt runoff based on the analysis of the advantages of meteorological elements. Background Art

[0002] Snowmelt runoff is an important source of runoff in cold and arid regions. For example, the snowmelt runoff in a certain source area is an important part of the runoff in a certain river basin, and the snowmelt runoff in a certain plateau area can account for 30% - 40% of the total annual runoff. In spring, if the temperature rises rapidly and the snow cover melts quickly, it is extremely easy to trigger spring floods. Therefore, understanding the contribution and proportion of snowmelt to runoff is of great significance for water resources management and flood control and disaster reduction. Currently, the main methods for calculating snowmelt runoff mainly use methods of runoff segmentation, such as electronic filtering method, isotope chemistry method, hydrological model method, graphical analysis method, energy balance method, and Eckhardt recursive digital filtering method, etc. However, due to the relatively complex composition of the runoff sources in a certain source area, the spring runoff composition may include rainfall runoff, snowmelt runoff, ice melt runoff, etc., and there are many factors affecting runoff, such as previous snowmelt, runoff, rainfall, evaporation, soil water content, etc., which will all contribute to the runoff volume to varying degrees. The currently commonly used methods are all direct processing of runoff data, without including the physical causes of the influence of meteorological and water resources elements on runoff composition, lacking in the mechanism and interpretability of runoff changes, and being unfavorable for grasping and predicting snowmelt runoff. Based on this, in order to quantitatively understand the contribution degree of each factor to runoff, especially the influence of snowmelt on runoff, and to achieve scientific management and efficient utilization of water resources, the present invention proposes a method for calculating the snowmelt runoff volume in a certain source area based on meteorological element analysis. Summary of the Invention

[0003] The purpose of the present invention is to provide a method for calculating snowmelt runoff based on the analysis of the advantages of meteorological elements in view of the deficiencies of the above-mentioned existing technologies.

[0004] The technical solution adopted by the present invention is as follows: A method for calculating snowmelt runoff based on the analysis of the advantages of meteorological elements, including the following specific steps: S1. Obtain the runoff data of the target hydrological station; S2. Based on the runoff data of the target hydrological station, construct a target influence monthly factor matrix; S3. Construct multiple models and calculate the contribution rate of each target influence factor to runoff in different models: S31. Establish a monthly regression equation subsystem model containing target influence factors based on the runoff data of the target hydrological station and the target influence monthly factor matrix; S32. Analyze the goodness of fit of each target influence factor to the monthly regression equation subsystem model based on the relative importance method, and calculate the average contribution increment of the goodness of fit; S33. Obtain the contribution degree of each target influencing factor to runoff in the monthly regression equation subsystem model based on the average contribution increment of the goodness of fit. S34. Based on the runoff of the target hydrological station and the target influencing monthly factor matrix, construct multiple target machine learning models. S35. Evaluate the multiple target machine learning models to obtain the goodness of fit of the multiple target machine learning models. S36. Based on the goodness of fit of the multiple target machine learning models, calculate the contribution degree of each target influencing factor to runoff in the multiple target machine learning models through the model interpretability algorithm. S37. Based on the fitting effects of the monthly regression equation subsystem model and the multiple target machine learning models, assign corresponding weights to different models. S38. Based on the assigned weights, calculate the contribution rate of each target influencing factor to runoff in different models. S4. Calculate the snowmelt runoff.

[0005] Furthermore, S1 specifically includes the following steps: S11. Determine the target area and the hydrological stations in the target area. S12. Select a certain hydrological station in the target area as the target hydrological station. S13. Select the long-term monthly runoff data of the target hydrological station as the runoff data of the target hydrological station: Let the year in the long-term monthly runoff data be , and the month be , then the runoff data of the target hydrological station is , where , = 1, 2, 3..., 12, .

[0006] Furthermore, S2 specifically includes the following steps: S21. Based on the runoff data of the target hydrological station, select the meteorological elements affecting the runoff of the target hydrological station as the target influencing factors. S22. Arrange the target influencing factors in the order of the corresponding years and months of the runoff data of the target hydrological station, and construct the target influencing factor dataset. S23. Based on the target influencing factor dataset, use the partial correlation method to calculate the leading and lagging partial correlation coefficients between the runoff data of the target hydrological station and each target influencing factor, and construct the partial correlation coefficient tensor. S24. Based on the partial correlation coefficient tensor, find the month corresponding to the maximum partial correlation coefficient between each target influencing factor and the runoff of the target hydrological station to obtain the target influencing monthly factors of the runoff of the target hydrological station. S25. Obtain the monthly runoff maximum partial correlation coefficient matrix based on the month in S24; S26. Construct the target impact month factor matrix based on the target impact month factor and the monthly runoff maximum partial correlation coefficient matrix.

[0007] Furthermore, the target impact factor dataset is as follows: (1) In formula (1), is the total number of target impact factor types; is the year corresponding to the target impact factor, is the month corresponding to the target impact factor, where , corresponds to the values in S13; The partial correlation coefficient tensor is as follows: (2) In formula (2), is the correlation relationship between the monthly runoff data of the target hydrological station from January to December and different months corresponding to the target impact factors; The maximum partial correlation coefficient matrix is as follows: (3) In formula (3), is the maximum monthly partial correlation coefficient of the th target impact factor affecting the runoff in the th month; Based on the target impact month factor and the monthly runoff maximum partial correlation coefficient matrix , construct the target impact month factor matrix .

[0008] Furthermore, in the monthly regression equation sub - model, the number of sub - models for each month is 2 n-1 pieces, and a total of 2 n-1 × sub - models can be established. Among them, the regression equation for a certain month can be expressed as: (4) The goodness of fit can be expressed as: (5) In formula (5), TSS is the total sum of squared deviations; RSS is the regression sum of squares;ESS is the sum of squared residuals; The contribution of each target influencing factor to runoff in the monthly regression equation subsystem model is as follows: (6).

[0009] Furthermore, the construction process of the multiple target machine learning models specifically includes the following steps: 1) Use XGboost, LSTM or random forest intelligent prediction methods to construct types of machine learning models as follows: ① Divide the runoff of the target hydrological station and the target influencing monthly factor matrix into a training set and a test set ; ② Train each machine learning model with the training set to obtain multiple target machine learning models , and input the test set into the multiple target machine learning models to obtain the prediction results of the runoff data by the types of machine learning models; 2) Evaluate the types of target machine learning models to obtain the goodness of fit of the multiple target machine learning models ; ; 3) Calculate the contribution of each target influencing factor to runoff in the types of target machine learning models through the model interpretability algorithm as follows: as follows: (7) 4) Based on the goodness of fit of the monthly regression equation subsystem model and the goodness of fit of the multiple target machine learning models , obtain the fitting effects of the ( ) types of models ; 5) Based on the fitting effects of the ( ) types of models , assign different weights to the ( ) types of models to obtain the corresponding weights : (8) In formula (7), the type of model, represents the proportion of the fitting effect of a certain model in the sum of the fitting effects of all models; 6) Based on the weights , the contribution rate of each target impact factor to runoff, that is, the weighted average final contribution rate of each target impact factor, is calculated. , and the specific calculation is as follows: (9).

[0010] Further, S4 specifically includes the following steps: S41. Based on the contribution rate of each target impact factor to runoff in different models, a contribution rate matrix of impact factors for the runoff of the target hydrological station is constructed. S42. The impact factors related to snowmelt are screened from the target impact factors as the target snowmelt impact factors. S43. The sum of the contribution rates corresponding to the target snowmelt impact factors is multiplied by the runoff data of the target hydrological station to obtain the runoff data related to snowmelt, that is, the snowmelt runoff.

[0011] Further, the contribution rate matrix of impact factors for the runoff of the target hydrological station , is specifically as follows: (10).

[0012] Further, the snowmelt runoff , and the specific calculation is as follows: (11) In formula (11), is the year corresponding to the target impact factor, is the month corresponding to the target impact factor, is the number of target snowmelt impact factors.

[0013] The present invention has the following beneficial effects compared with the prior art: The present invention fully considers the complexity of runoff composition, uses the partial correlation method to eliminate the influence of other impact factors on the runoff data of the target hydrological station, and considers different impact factors for the runoff in different months; considering the fitting effects of different models and the contribution rates of each target impact factor to runoff in different models, a multi-model group of monthly regression equation sub-models and various target machine learning models is constructed, and a weighted average multi-model integrated contribution rate calculation method is proposed based on the fitting effects of different models, effectively overcoming the non-universality of the contribution rate of impact factors in a single model, and enhancing the scientificity and interpretability in the current calculation method of snowmelt runoff. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present invention will be further described below in conjunction with the drawings and specific embodiments: Figure 1 is a schematic flow chart of a snowmelt runoff calculation method based on the advantage analysis of meteorological elements according to the present invention. Specific implementation mode

[0015] Embodiment As Figure 1 shown, a snowmelt runoff calculation method based on meteorological element advantage analysis includes the following specific steps: S1. Obtain the runoff data of the target hydrological station; Specifically, S11. Determine the target area and the hydrological stations in the target area; S12. Select a certain hydrological station in the target area as the target hydrological station; S13. Select the long-term monthly runoff data of the target hydrological station as the runoff data of the target hydrological station: Let the year in the long-term monthly runoff data be , and the month be , then the runoff data of the target hydrological station is , where , = 1, 2, 3..., 12; To ensure sufficient sample size, it is necessary to meet the sample size requirement of , that is, at least 30 years of runoff data of the target hydrological station are taken to calculate the snowmelt runoff for all months of the year; In this embodiment, the target area is determined as a certain river basin and multiple hydrological stations in the river basin. A certain station in the river basin is selected as the target hydrological station, and the monthly runoff data from 1979 to 2019 are selected as the runoff data of the station. The sample size is 41 years from 1979 to 2019, meeting the requirements; S2. Based on the runoff data of the target hydrological station, construct a target impact monthly factor matrix: Specifically, S21. Based on the runoff data of the target hydrological station, select the meteorological elements that affect the runoff of the target hydrological station as the target impact factors; S22. Arrange the target impact factors in the order of the corresponding years and months of the runoff data of the target hydrological station to construct a target impact factor data set; S23. Based on the target impact factor data set, use the partial correlation method to calculate the leading and lagging partial correlation coefficients between the runoff data of the target hydrological station and each target impact factor, and construct a partial correlation coefficient tensor; S24. Based on the partial correlation coefficient tensor, find the month corresponding to the maximum partial correlation coefficient between each target impact factor and the runoff of the target hydrological station to obtain the target impact monthly factor of the runoff of the target hydrological station; S25. Based on the months in S24, obtain the monthly runoff maximum partial correlation coefficient matrix; S26. Based on the target impact monthly factor and the monthly runoff maximum partial correlation coefficient matrix, construct a target impact monthly factor matrix;

[0016] In this embodiment, based on the obtained runoff data of the target hydrological station for 41 years , the meteorological elements affecting the runoff of the target hydrological station are selected as the target influencing factors, and the specific operations are as follows: Study 130 circulation indices of the National Climate Center, including 88 atmospheric circulation indices, 26 sea surface temperature indices, and 16 other types of climate indices. Select rainfall, runoff, snow depth, evaporation, and soil moisture content as the target influencing factors affecting the runoff of a certain station; To facilitate the calculation of the leading and lagging partial correlation coefficients between the runoff data of the target hydrological station and each target influencing factor, one year is taken before and after the runoff data of a certain station, that is, 43 years of monthly runoff data from 1978 to 2020 are selected for the above target influencing factors; Arrange the monthly runoff data corresponding to the target influencing factors in chronological order of year and month to construct a target influencing factor data set , specifically as follows: (1) In formula (1), is the total number of types of target influencing factors; is the year corresponding to the target influencing factor, is the month corresponding to the target influencing factor, where , corresponds to the values in S13; Based on the target influencing factor data set, the leading and lagging partial correlation coefficients between the runoff data of the target hydrological station and each target influencing factor are calculated using the partial correlation method to construct a partial correlation coefficient tensor , specifically as follows: (2) In formula (2), is the correlation relationship between the monthly runoff data of the target hydrological station from January to December and different months corresponding to the target influencing factors; the use of the partial correlation method for calculation is to avoid the influence of the interaction between multiple factors on the correlation; in addition, the calculated partial correlation coefficients can be subjected to a significance test to ensure the reliability and effectiveness of the results; Based on the partial correlation coefficient tensor , find the month corresponding to the maximum partial correlation coefficient between each target influencing factor and the runoff of the target hydrological station to obtain the target influencing month factor of the runoff of the target hydrological station; By analyzing the leading-lagging partial correlation coefficient results between the runoff data of the target hydrological station and each target influencing factor, it is found that in this embodiment, the month corresponding to the maximum partial correlation coefficient between the selected target influencing factors and the runoff of the target hydrological station is the same period or the previous month. That is, these target influencing factors have a significant impact on the runoff of a certain station in the same period or one month in advance. Therefore, the target influencing factors corresponding to the same period and the previous month are considered as the target influencing monthly factors of the runoff of a certain station, as shown in Table 1; the target influencing factors in other months are omitted because their impact on the runoff of a certain station is relatively small.

[0017] Table 1 shows the partial correlation coefficients between the monthly runoff of a certain station and the target influencing factors, specifically the partial correlation coefficients between the monthly runoff of a certain station and each target influencing factor in the same period or one month in advance: Table 1 Partial correlation coefficients between the monthly runoff of a certain station and the target influencing factors

[0018] In Table 1, ✽✽ indicates passing the 99% significance test, and ✽ indicates passing the 95% significance test, indicating that the results are reliable and valid; Based on the months mentioned above, the monthly maximum partial correlation coefficient matrix of runoff is obtained , specifically as follows: (3) In Equation (3), is the maximum monthly partial correlation coefficient of the th target influencing factor affecting the runoff in the th month; Based on the target influencing monthly factors and the monthly maximum partial correlation coefficient matrix of runoff , the target influencing monthly factor matrix is constructed .

[0019] S3. Calculate the contribution rate of the target influencing factor to the runoff: Specifically, S31. Based on the runoff data of the target hydrological station and the target influencing monthly factor matrix, a monthly regression equation subsystem model containing the target influencing factors is established; S32. Based on the relative importance method, analyze the goodness of fit of each target influencing factor to the monthly regression equation subsystem model, and calculate the average contribution increment of the goodness of fit; S33. Based on the average contribution increment of the goodness of fit, obtain the contribution degree of each target influencing factor to the runoff in the monthly regression equation subsystem model; S34. Based on the runoff of the target hydrological station and the target influencing monthly factor matrix, construct various target machine learning models; S35. Evaluate the various target machine learning models to obtain the goodness of fit of the various target machine learning models; S36. Calculate the contribution degree of each target influencing factor to runoff in multiple target machine learning models through the model interpretability algorithm based on the goodness of fit of multiple target machine learning models; S37. Assign corresponding weights to different models based on the fitting effects of the monthly regression equation subsystem models and multiple target machine learning models; S38. Calculate the contribution rate of each target influencing factor to runoff in different models based on the assigned weights; In this embodiment, in the established monthly regression equation subsystem model, the number of subsystem models for each month is 2 n-1 pieces, and a total of 2 n-1 × subsystem models can be established. Among them, the regression equation for a certain month can be expressed as: (4) Analyze the goodness of fit of each target influencing factor to the monthly regression equation subsystem model based on the relative importance method and calculate the average contribution increment of the goodness of fit , where the goodness of fit can be expressed as: (5) In formula (5), TSS is the total sum of squared deviations; RSS is the regression sum of squares; ESS is the residual sum of squares; Based on the average contribution increment of the goodness of fit , obtain the contribution degree of each target influencing factor to runoff in the monthly regression equation subsystem model, specifically as follows: (6) Construct machine learning models using XGboost, LSTM or random forest intelligent prediction methods , specifically as follows: ① Divide the runoff of the target hydrological station and the target influencing monthly factor matrix into a training set and a test set in proportion; ② Train each machine learning model with the training set to obtain multiple target machine learning models , and input the test set into the multiple target machine learning models to obtain the prediction results of the machine learning models for runoff data; For Evaluating multiple target machine learning models to obtain the goodness of fit of multiple target machine learning models ; Calculating through a model interpretability algorithm to obtain the contribution degree of each target influencing factor to runoff in multiple target machine learning models, specifically as follows: Specifically as follows: (7) Among them, the basic principle of the model interpretability algorithm is: calculating the Shapely value of each target influencing factor in each model, that is, calculating the weighted average marginal contribution when a target influencing factor is added to multiple target machine learning models, and then obtaining the contribution degree of the target influencing factor in multiple target machine learning models; Based on the goodness of fit of the sub-model of the monthly regression equation system and the goodness of fit of multiple target machine learning models , obtaining the fitting effect of ( ) models ; Based on the fitting effect of ( ) models , assigning different weights to ( ) models to obtain the corresponding weights : (8) In formula (7), model type, represents the proportion of the fitting effect of a certain model in the sum of the fitting effects of all models; Based on the weight , calculating the contribution rate of each target influencing factor to runoff, that is, the weighted average final contribution rate of each target influencing factor , specifically calculated as follows: (9);

[0020] S4. Calculating the snowmelt runoff: Specifically, S41. Based on the contribution rate of each target influencing factor to runoff in different models, constructing a contribution rate matrix of runoff influencing factors for the target hydrological station; S42. Screening the influencing factors related to snowmelt from the target influencing factors as the target snowmelt influencing factors; S43. Multiplying the sum of the contribution rates corresponding to the target snowmelt influencing factors by the runoff data of the target hydrological station to obtain the runoff data related to snowmelt, that is, the snowmelt runoff; In this embodiment, based on ( Weighted average final contribution rate of each target influencing factor to construct months Contribution rate matrix of runoff influencing factors of the target hydrological station for various target influencing factors as follows: (10) Screen out the influencing factors related to snowmelt from the target influencing factors as the target snowmelt influencing factors; Multiply the sum of the contribution rates corresponding to the target snowmelt influencing factors by the runoff data of the target hydrological station to obtain the runoff data related to snowmelt, i.e., snowmelt runoff The specific calculation is as follows: (11) In formula (11), in formula (11), is the year corresponding to the target influencing factor, is the month corresponding to the target influencing factor, is the number of target snowmelt influencing factors; As shown in Table 2, it shows the contribution rates of runoff influencing factors for each month of a certain station; when calculating the snowmelt runoff, the sum of the contribution rates of the previous month's snow cover and the current month's snow cover is selected as the snowmelt contribution rate, as follows: Table 2 Contribution rates of runoff influencing factors for each month of a certain station

[0021] The present invention fully considers the complexity of runoff composition, uses the partial correlation method to eliminate the influence of other influencing factors on the runoff data of the target hydrological station, and considers different influencing factors for the runoff in different months; considering the fitting effects of different models and the contribution rates of each target influencing factor to runoff in different models, a multi-model group of monthly regression equation sub-models and various target machine learning models is constructed, and a weighted average multi-model integration contribution rate calculation method is proposed based on the fitting effects of different models, effectively overcoming the non-universality of the contribution rate of influencing factors in a single model, and enhancing the scientificity and interpretability in the current snowmelt runoff calculation method.

[0022] The embodiments described above are only used to describe the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the principles and essence of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for calculating snowmelt runoff based on the analysis of meteorological element advantages, characterized in that, It includes the following steps specifically: S1. Obtain the runoff data of the target hydrological station; S2. Based on the runoff data of the target hydrological station, construct the target impact monthly factor matrix; S3. Construct multiple models and calculate the contribution rate of each target impact factor to runoff in different models: S31. Establish a monthly regression equation subsystem model containing target impact factors based on the runoff data of the target hydrological station and the target impact monthly factor matrix; S32. Analyze the goodness of fit of each target impact factor to the monthly regression equation subsystem model based on the relative importance method, and calculate the average contribution increment of the goodness of fit; S33. Based on the average contribution increment of the goodness of fit, obtain the contribution degree of each target impact factor to runoff in the monthly regression equation subsystem model; S34. Based on the runoff of the target hydrological station and the target impact monthly factor matrix, construct multiple target machine learning models; S35. Evaluate the models of multiple target machine learning models to obtain the goodness of fit of multiple target machine learning models; S36. Based on the goodness of fit of multiple target machine learning models, calculate the contribution degree of each target impact factor to runoff in multiple target machine learning models through the model interpretability algorithm; S37. Based on the fitting effects of the monthly regression equation subsystem model and multiple target machine learning models, assign corresponding weights to different models; S38. Based on the assigned weights, calculate the contribution rate of each target impact factor to runoff in different models; S4. Calculate the snowmelt runoff.

2. The snowmelt runoff calculation method based on the analysis of meteorological element advantages according to claim 1, characterized in that, S1 specifically includes the following steps: S11. Determine the target area and the hydrological stations in the target area; S12. Select a certain hydrological station in the target area as the target hydrological station; S13. Select the long-term monthly runoff data of the target hydrological station as the runoff data of the target hydrological station: Let the year in the long - term monthly runoff data be , the month be , then the runoff data of the target hydrological station is , where , = 1, 2, 3..., 12, .

3. A snowmelt runoff calculation method based on the analysis of meteorological element advantages according to claim 1, characterized in that, S2 specifically includes the following steps: S21. Based on the runoff data of the target hydrological station, select the meteorological elements affecting the runoff of the target hydrological station as the target impact factors; S22. Arrange the target impact factors in the order of the corresponding years and months of the runoff data of the target hydrological station, and construct the target impact factor data set; S23. Based on the target impact factor data set, use the partial correlation method to calculate the lead-lag partial correlation coefficients between the runoff data of the target hydrological station and each target impact factor, and construct the partial correlation coefficient tensor; S24. Based on the partial correlation coefficient tensor, find the month corresponding to the maximum partial correlation coefficient between each target impact factor and the runoff of the target hydrological station to obtain the target impact monthly factor of the runoff of the target hydrological station; S25. Obtain the monthly runoff maximum partial correlation coefficient matrix based on the months in S24; S26. Based on the target impact monthly factor and the monthly runoff maximum partial correlation coefficient matrix, construct the target impact monthly factor matrix.

4. The snowmelt runoff calculation method based on meteorological element advantage analysis according to claim 3, wherein The target impact factor dataset , which is specifically as follows: （1） In formula (1), is the total number of types of target impact factors; is the year corresponding to the target impact factor, is the month corresponding to the target impact factor, where and correspond to the values in S13; The partial correlation coefficient tensor , is specifically as follows: （2） In formula (2), is the correlation relationship between the monthly runoff data of the target hydrological station from January to December and the different months corresponding to the target influencing factors; The maximum partial correlation coefficient matrix , is specifically as follows: （3） In formula (3), is the maximum monthly partial correlation coefficient of the th target influencing factor affecting the monthly runoff in the th month; Based on the target impact monthly factor and the maximum partial correlation coefficient matrix of monthly runoff , the target impact monthly factor matrix is constructed .

5. A snowmelt runoff calculation method based on the analysis of meteorological element advantages according to claim 1, characterized in that, In the monthly regression equation sub-model, the number of sub-models for each month is 2 n-1 , and a total of 2 n-1 × sub-models can be established. Among them, the regression equation system for a certain month can be expressed as: （4） The goodness of fit can be expressed as: （5） In formula (5), TSS is the total sum of squares of deviations; RSS is the regression sum of squares; ESS is the residual sum of squares; The contribution degree of each target impact factor to runoff in the monthly regression equation sub-model , is specifically as follows: （6）。 6. The snowmelt runoff calculation method based on the analysis of meteorological element advantages according to claim 1, wherein The construction process of the multiple target machine learning models specifically includes the following steps: 1) Construct machine learning models using XGboost, LSTM or random forest intelligent prediction methods, specifically as follows: ①Divide the runoff of the target hydrological station and the target impact monthly factor matrix proportionally into a training set and a test set ; ②Train each machine learning model with the training set to obtain multiple target machine learning models , and input the test set into multiple target machine learning models to obtain the prediction results of runoff data by several machine learning models; 2) For the target machine learning models, perform model evaluation to obtain the goodness of fit of the multiple target machine learning models ; 3) Calculate, through the model interpretability algorithm, the contribution degree of each target impact factor to runoff in target learning models, as follows: Specifically: （7） 4) Goodness of fit of the monthly regression equation sub-model and the goodness of fit of multiple target machine learning models , obtaining the goodness-of-fit of ([[]] ) models ; 5) Based on ( ) kinds of fitting effects of the model , different weights are assigned to ( ) kinds of models to obtain the corresponding weights : （8） In formula (7), Model type, represents the proportion of the fitting effect of a certain model in the sum of the fitting effects of all models; 6) Based on weights , calculate the contribution rate of each target impact factor to runoff, that is, the weighted average final contribution rate of each target impact factor , and the specific calculation is as follows: （9）。 7. A snowmelt runoff calculation method based on the analysis of meteorological element advantages according to claim 1, characterized in that S4 specifically includes the following steps: S41. Based on the contribution rate of each target impact factor to runoff in different models, construct the target hydrological station runoff impact factor contribution rate matrix; S42. Screen out the impact factors related to snowmelt from the target impact factors as the target snowmelt impact factors; S43. Multiply the sum of the contribution rates corresponding to the target snowmelt impact factors by the runoff data of the target hydrological station to obtain the runoff data related to snowmelt, that is, the snowmelt runoff volume.

8. A method for calculating snowmelt runoff based on the analysis of meteorological element advantages according to claim 7, characterized in that, The contribution rate matrix of runoff influencing factors of the target hydrological station , which is specifically as follows: （10）。 9. A method for calculating snowmelt runoff based on the analysis of meteorological element advantages according to claim 7, characterized in that, The snowmelt runoff , and the specific calculation is as follows: （11） In formula (11), is the year corresponding to the target impact factor, is the month corresponding to the target impact factor, is the number of the target snowmelt impact factors.

Images