Progressive strategy-based Yangtze River midstream plum emergence time prediction method

Abstract

The invention provides a progressive strategy-based Yangtze River midstream plum emergence time prediction method. The method comprises the following steps of: selecting a prediction time period and a Yangtze River midstream representative partition suitable for calculation; evaluating the relevance of multiple climatic factors according to a prediction object, and selecting a key climatic factor set which affects the plum output time; calculating meteorological factors corresponding to different time scenes and space scenes by adopting a time separation and space layering method, and constructing a multi-layer meteorological factor index library; and according to the key climate factor set and the multilayer meteorological factor index library, adopting a long-medium-short term progressive strategy to realize prediction of the mid-swimming plum emergence time of the Yangtze River. According to the method, a key climate factor set and a multi-layer meteorological factor index library which influence the mid-run plum emergence time of the Yangtze River can be established, and seamless full coverage of the prediction period can be realized by using a long-medium-short term progressive strategy, so that a new reference is provided for the method for predicting the mid-run plum emergence time of the Yangtze River, and business application and popularization are facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of long-term prediction, and in particular to a method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a gradual strategy. Background Art

[0002] The meteorological department's current plum rain indicators divide the Jianghuai River Basin's plum rain season into three monitoring zones: the Jiangnan region, the middle and lower Yangtze River region, and the Jianghuai region. This defined plum rain season covers a relatively small area. Specifically for the Yangtze River Basin, the middle and lower Yangtze River encompasses the aforementioned plum rain zones, with the two lake systems being the primary runoff-producing areas. However, using this standard would result in widespread heavy rainfall in the two lake systems, while weak rainfall in the middle and lower Yangtze River mainstream would be considered non-plum rain events, hindering flood control decision-making. Current plum rain season operations primarily rely on real-time monitoring to determine the end of the plum rain season. As this service evolves, there is an urgent need to integrate real-time monitoring with forecasts to determine the end date of the plum rain season. The current forecast method for the end of the plum rain season is mainly based on the short- and medium-term forecast perspective. It is based on the standards of the current meteorological department and uses short- and medium-term numerical model data forecasts. The forecast period is generally short. Based on this, in order to further extend the forecast period for judging the end of the plum rain season and reserve a longer forecast period and response time for the adjustment of reservoir water storage strategies, power generation situation judgment, flood and drought prevention, water resources allocation and other work, the present invention proposes a method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a progressive strategy. Summary of the Invention

[0003] The purpose of the present invention is to address the deficiencies of the above-mentioned prior art and provide a method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a progressive strategy.

[0004] To achieve the above object, the present invention adopts the following technical solutions:

[0005] The present invention provides a method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a progressive strategy, comprising the following steps:

[0006] S1. Determine the forecast object: select the forecast period and representative sub-regions of the middle reaches of the Yangtze River, and select the long-term series data of the end of the plum rain season;

[0007] S2. Determination of a set of highly correlated key climate factors: Based on the time series of the end of the plum rain season in representative subregions, the correlation between multiple climate factors and the end of the plum rain season is evaluated, key climate factors with high correlation with the prediction object are determined, and a set of key climate factors is constructed;

[0008] S3. Determination of a multi-layer meteorological factor index library: Using the time separation method, different time periods are divided based on different plum rain season scenarios; using the spatial stratification method, meteorological factors corresponding to multiple spatial scenarios under different time scenarios are calculated to construct a multi-layer meteorological factor index library;

[0009] S4. Prediction of the end of the plum rain season: Based on the key climate factor set and the multi-layer meteorological factor index library obtained in S2 and S3, a progressive analysis method from long-term climate factors to medium- and short-term meteorological factors is adopted to predict the end of the plum rain season in the middle reaches of the Yangtze River.

[0010] Furthermore, the specific steps of S1 are:

[0011] For each representative subregion in the middle reaches of the Yangtze River, long-term series data of the end of the plum rain season are selected, where the data year is y, the month is m, and the specific date is i. The prediction sample of the end of the plum rain season is T(m,i,y), where m=1,2,3,..,12, i is a positive integer less than 31, and y∈[y i ,y j ], in order to ensure sufficient sample size, (y j -y i )≥50, at least 50 years of samples are taken.

[0012] Furthermore, the S2 is specifically:

[0013] S21. Set the sample data of the climate factors used for the prediction of the end of the plum rain season to C(k,n,z), where k is the type of factor, n is the month of the factor, and Z is the year corresponding to the factor sample;

[0014] S22. Since the plum rain season occurs in summer, considering the stability of the climate state, the winter climate factors are used to predict the end of the plum rain season in summer. In C(k,n,z), n∈[n i ,n j ], m∈[m i ,m j ], then (m j -n i )≤6, that is, the difference in months between the forecast object and the forecast element cannot exceed 6 months;

[0015] S23. Find the correlation coefficients between each climate factor sample and the plum rain season time series to obtain the correlation matrix R zjn :

[0016]

[0017] Among them, z is different climate factors; j is the end of the plum rain season; R zjn The correlation between the end of the plum rain season and n influencing factors in different months is calculated, and the significance test of the correlation coefficient is performed;

[0018] S24. Based on the sample data and the correlation coefficient matrix, select climate factors with high correlation coefficients and passing the significance test, and finally obtain the key climate factor matrix with high correlation coefficients:

[0019]

[0020] Among them, n in is the correlation coefficient, It is the climate factor matrix with high correlation coefficient corresponding to the sample T(m,i,y) of the prediction of the end of plum rain season.

[0021] Furthermore, the S3 is specifically:

[0022] S31. Use the time separation TMA method to perform time separation on the samples T(m,i,y) for prediction of the end of the plum rain season. Classify the samples according to different time scene characteristics. Set the number of time scenes to λ, and the calculation is:

[0023]

[0024] Among them, T early (m,i,y) is the sample with early end of plum rain season, TMA(T early ,α) is the matrix of the samples T(m,i,y) predicted by the end of the plum rain season after time separation, and the number of all samples with early end of the plum rain season is α i ;

[0025]

[0026] Among them, T late (m,i,y) is the sample with late end of plum rain season, TMA(T late ,b) is the matrix of the predicted samples T(m,i,y) after time separation and classification, and the number of all samples with late plum rain end time is b i ;

[0027] Get the final scene sum:

[0028] α i +b j =λ;

[0029] S32, using PMA method to analyze the matrix TMA (T early ,α) and the matrix TMA (T late ,b) Perform spatial stratification respectively, classify the samples according to the different height characteristics of the circulation field, and set the number of spatial stratification scenes to β, then the calculation formula is:

[0030]

[0031] Among them, PMA(TMA(T early ,α),e) is the predicted sample TMA (T early,α) After the spatial stratification, the total number of spatial scenes is e:

[0032]

[0033] Among them, PMA(TMA(T late ,b) and f) are the predicted samples TMA (T late ,b) The matrix after spatial stratification, the total number of spatial scenes is f:

[0034] The final scene sum is:

[0035] e i +f j =β;

[0036] S33. Use the methods in S31 and S32 to establish a multi-layer meteorological factor index matrix PMA (TMA,o).

[0037] Furthermore, the S4 is specifically:

[0038] S41. According to the correlation rate of different climate factors, different weights ω are assigned to different climate factors:

[0039]

[0040] According to the different weights of different climate factors, the weighted average final contribution rate θ of each key climate factor is calculated:

[0041]

[0042] Set the trigger conditions of climate factors, θ i ≥0.7 or θ i When the temperature is ≤-0.7, the corresponding climate scene of the end of the plum rain season will be automatically triggered;

[0043] S42. Based on the meteorological factor indicator matrix PMA (TMA, e), meteorological factor triggering conditions are set under different scenarios. When the meteorological factors meet the requirements of a specific scenario, the corresponding end of the plum rain season date can be predicted.

[0044] The beneficial effects of the present invention are as follows: a new method for predicting the end of the plum rain season is proposed from the perspective of a gradual decision-making based on the long-term, medium-term and short-term, which fully considers the complexity of the climate background and atmospheric circulation, and selects a set of key climate factors that affect the end of the plum rain season by evaluating the correlation between multiple climate factors. By adopting the time separation and space stratification method, the meteorological factors corresponding to different time scenes and spatial scenes are considered to construct a multi-layer meteorological factor index library. Based on the key climate factor set and the multi-layer meteorological factor index library, the long-term, medium-term and short-term gradual forecasting situation is adopted to realize the prediction of the end of the plum rain season in the middle reaches of the Yangtze River. It can achieve seamless full coverage of different forecast forecast periods, which is conducive to further extending the forecast period and improving the forecast accuracy, and provides a new reference for the prediction method of the end of the plum rain season in the middle reaches of the Yangtze River. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flowchart of a method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a gradual strategy;

[0046] Figure 2 It is a physical conceptual model of summer precipitation in the middle and lower reaches of the Yangtze River;

[0047] Figure 3 It is an indicator of the circulation during the end of the plum rain season in the middle and lower reaches of the Yangtze River. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0049] See also Figure 1 ,A method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on ,gradual strategy, includes the following steps;

[0050] S1. Determine the forecast object: select the forecast period and representative sub-regions of the middle reaches of the Yangtze River, and select the long-term series data of the end of the plum rain season;

[0051] S2. Determination of a set of highly correlated key climate factors: Based on the time series of the end of the plum rain season in representative subregions, the correlation between multiple climate factors and the end of the plum rain season is evaluated, key climate factors with high correlation with the prediction object are determined, and a set of key climate factors is constructed;

[0052] S3. Determination of a multi-layer meteorological factor index library: Using the time separation method, different time periods are divided based on different plum rain season scenarios; using the spatial stratification method, meteorological factors corresponding to multiple spatial scenarios under different time scenarios are calculated to construct a multi-layer meteorological factor index library;

[0053] S4. Prediction of the end of the plum rain season: Based on the key climate factor set and the multi-layer meteorological factor index library obtained in S2 and S3, a progressive analysis method from long-term climate factors to medium- and short-term meteorological factors is adopted to predict the end of the plum rain season in the middle reaches of the Yangtze River.

[0054] The specific steps of S1 are:

[0055] For each representative subregion in the middle reaches of the Yangtze River, long-term series data of the end of the plum rain season are selected, where the data year is y, the month is m, and the specific date is i. The prediction sample of the end of the plum rain season is T(m,i,y), where m=1,2,3,..,12, i is a positive integer less than 31, and y∈[y i ,y j ], in order to ensure sufficient sample size, (y j -y i )≥50, at least 50 years of samples are taken.

[0056] In this embodiment, the Dongting Lake water system in the middle reaches of the Yangtze River and the main stream of the middle reaches of the Yangtze River are selected as the forecast objects. The daily rainfall data from 1961 to 2023 are selected. The factor sample length is 63 years from 1961 to 2023, which meets the sample size requirement.

[0057] The S2 is specifically:

[0058] S21. Set the sample data of the climate factors used for the prediction of the end of the plum rain season to C(k,n,z), where k is the type of factor, n is the month of the factor, and Z is the year corresponding to the factor sample;

[0059] S22. Since the plum rain season occurs in summer, considering the stability of the climate state, the winter climate factors are used to predict the end of the plum rain season in summer. In C(k,n,z), n∈[n i ,n j ], m∈[m i ,m j ], then (m j -n i )≤6, that is, the difference in months between the forecast object and the forecast element cannot exceed 6 months;

[0060] In this embodiment, the sea surface temperature, snow cover on the Qinghai-Tibet Plateau, sea ice and 130 monthly circulation indices from the National Climate Center for a total of 63 years from 1961 to 2023 are studied.

[0061] S23. Find the correlation coefficients between each climate factor sample and the plum rain season time series to obtain the correlation matrix R zjn :

[0062]

[0063] Among them, z is different climate factors; j is the end of the plum rain season; R zjn The correlation between the end of the plum rain season and n influencing factors in different months is calculated, and the significance test of the correlation coefficient is performed;

[0064] S24. Based on the sample data and the correlation coefficient matrix, select climate factors with high correlation coefficients and passing the significance test, and finally obtain the key climate factor matrix with high correlation coefficients:

[0065]

[0066] Among them, n in is the correlation coefficient, It is the climate factor matrix with high correlation coefficient corresponding to the sample T(m,i,y) of the prediction of the end of plum rain season.

[0067] In this embodiment, key climate factors highly correlated with the plum rain season and the end of the plum rain season are found, including sea temperature, snow cover, and sea ice.

[0068] The S3 is specifically:

[0069] S31. Use the time separation TMA method to perform time separation on the samples T(m,i,y) for prediction of the end of the plum rain season. Classify the samples according to different time scene characteristics. Set the number of time scenes to λ, and the calculation is:

[0070]

[0071] Among them, T early (m,i,y) is the sample with early end of plum rain season, TMA(T early ,α) is the matrix of the samples T(m,i,y) predicted by the end of the plum rain season after time separation, and the number of all samples with early end of the plum rain season is α i ;

[0072]

[0073] Among them, T late (m,i,y) is the sample with late end of plum rain season, TMA(T late ,b) is the matrix of the predicted samples T(m,i,y) after time separation and classification, and the number of all samples with late plum rain end time is b i ;

[0074] Get the final scene sum:

[0075] α i +b j =λ;

[0076] S32, using PMA method to analyze the matrix TMA (T early ,α) and the matrix TMA (T late ,b) Perform spatial stratification respectively, classify the samples according to the different height characteristics of the circulation field, and set the number of spatial stratification scenes to β, then the calculation formula is:

[0077]

[0078] Among them, PMA(TMA(T early ,α),e) is the predicted sample TMA (T early ,α) After the spatial stratification, the total number of spatial scenes is e:

[0079]

[0080] Among them, PMA(TMA(T late ,b) and f) are the predicted samples TMA (T late ,b) The matrix after spatial stratification, the total number of spatial scenes is f:

[0081] The final scene sum is:

[0082] e i +f j =β;

[0083] S33. Use the methods in S31 and S32 to establish a multi-layer meteorological factor index matrix PMA (TMA,o).

[0084] See also Figure 2 and Figure 3 In this embodiment, the main meteorological factors for the end of the plum rain season are found. ① The subtropical high pressure ridge jumps north to 25°N and north. ② The middle and lower reaches of the Yangtze River are controlled by the subtropical high pressure, or the main stream of the middle and lower reaches of the Yangtze River and its south are controlled by the subtropical high pressure, that is, the middle and lower reaches of the Yangtze River are controlled by the anticyclonic wind field. ③ The high-latitude meridional resistance high or trough ridge collapses or moves eastward, and the high-latitude westerly belt transforms into a relatively straight zonal circulation. ④ The shear convergence in the middle and lower reaches of the Yangtze River disappears in the middle and low layers, and turns into southwesterly winds, southerly winds, or southeasterly winds. ⑤ The intensity of the subtropical high pressure is significantly enhanced. ⑥ The mid-level water vapor transport shifts from the Bay of Bengal to the South China Sea and the western Pacific Ocean. ⑦ The middle and lower reaches of the Yangtze River change from a cyclonic circulation to an anticyclonic circulation in the difference field.

[0085] The S4 is specifically:

[0086] S41. According to the correlation rate of different climate factors, different weights ω are assigned to different climate factors:

[0087]

[0088] According to the different weights of different climate factors, the weighted average final contribution rate θ of each key climate factor is calculated:

[0089]

[0090] Set the trigger conditions of climate factors, θ i ≥0.7 or θ i When the temperature is ≤-0.7, the corresponding climate scene of the end of the plum rain season will be automatically triggered;

[0091] S42. Based on the meteorological factor indicator matrix PMA (TMA, e), meteorological factor triggering conditions are set under different scenarios. When the meteorological factors meet the requirements of a specific scenario, the corresponding end of the plum rain season date can be predicted.

[0092] In this embodiment, in the actual business forecast for 2024, it is predicted from the perspective of climate factors that the end of the plum rain season in 2024 will be late, and a more qualitative prediction is given. The specific end of the plum rain season is predicted from the perspective of meteorological factors. The actual situation is relatively consistent with the forecast, which provides an important reference for flood control decisions in 2024.

[0093] In summary, the present invention proposes a method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a progressive strategy, which fully considers the complexity of the climate background and atmospheric circulation, and selects a set of key climate factors that affect the end of the plum rain season by evaluating the correlation between multiple climate factors. By adopting the time separation and spatial stratification method, the meteorological factors corresponding to different time scenes and spatial scenes are considered to construct a multi-layer meteorological factor index library. Based on the key climate factor set and the multi-layer meteorological factor index library, the long-medium-short-term progressive forecast situation is adopted to realize the prediction of the end of the plum rain season in the middle reaches of the Yangtze River. It can achieve seamless full coverage of different forecast forecast periods, which is conducive to further extending the forecast period and improving the forecast accuracy, and provides a new reference for the prediction method of the end of the plum rain season in the middle reaches of the Yangtze River.

[0094] The above-described embodiments merely illustrate the implementation methods of the present invention. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a gradual strategy, characterized in that: The method includes the following steps: S1. Determine the forecast object: select the forecast period and representative sub-regions of the middle reaches of the Yangtze River, and select the long-term series data of the end of the plum rain season; S2. Determination of a set of highly correlated key climate factors: Based on the time series of the end of the plum rain season in representative subregions, the correlation between multiple climate factors and the end of the plum rain season is evaluated, key climate factors with high correlation with the prediction object are determined, and a set of key climate factors is constructed; S3. Determination of a multi-layer meteorological factor index library: Using the time separation method, different time periods are divided based on different plum rain season scenarios; using the spatial stratification method, meteorological factors corresponding to multiple spatial scenarios under different time scenarios are calculated to construct a multi-layer meteorological factor index library; S4. Prediction of the end of the plum rain season: Based on the key climate factor set and the multi-layer meteorological factor index library obtained in S2 and S3, a progressive analysis method from long-term climate factors to medium- and short-term meteorological factors is adopted to predict the end of the plum rain season in the middle reaches of the Yangtze River.

2. The method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a progressive strategy according to claim 1, characterized in that: The specific steps of S1 are: For each representative subregion in the middle reaches of the Yangtze River, long-term series data of the end of the plum rain season are selected, where the data year is y, the month is m, and the specific date is i. The prediction sample of the end of the plum rain season is T(m,i,y), where m=1,2,3,..,12, i is a positive integer less than 31, and y∈[y i ,y j ], in order to ensure sufficient sample size, (y j -y i )≥50, at least 50 years of samples are taken.

3. The method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a progressive strategy according to claim 2, characterized in that: The S2 is specifically: S21. Set the sample data of the climate factors used for the prediction of the end of the plum rain season to C(k,n,z), where k is the type of factor, n is the month of the factor, and Z is the year corresponding to the factor sample; S22. Since the plum rain season occurs in summer, considering the stability of the climate state, the winter climate factors are used to predict the end of the plum rain season in summer. In C(k,n,z), n∈[n i ,n j ], m∈[m i ,m j ], then (m j -n i )≤6, that is, the difference in months between the forecast object and the forecast element cannot exceed 6 months; S23. Find the correlation coefficients between each climate factor sample and the plum rain season time series to obtain the correlation matrix R zjn : Among them, z is different climate factors; j is the end of the plum rain season; R zjn The correlation between the end of the plum rain season and n influencing factors in different months is calculated, and the significance test of the correlation coefficient is performed; S24. Based on the sample data and the correlation coefficient matrix, select climate factors with high correlation coefficients and passing the significance test, and finally obtain the key climate factor matrix with high correlation coefficients: Among them, n in is the correlation coefficient, It is the climate factor matrix with high correlation coefficient corresponding to the sample T(m,i,y) of the prediction of the end of plum rain season.

4. The method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a progressive strategy according to claim 3, characterized in that: The S3 is specifically: S31. Use the time separation TMA method to perform time separation on the samples T(m,i,y) for prediction of the end of the plum rain season. Classify the samples according to different time scene characteristics. Set the number of time scenes to λ, and the calculation is: Among them, T early (m,i,y) is the sample with early end of plum rain season, TMA(T early ,α) is the matrix of the samples T(m,i,y) predicted to be early after time separation, and the number of samples with early time is α i ; Among them, T late (m,i,y) is the sample with late end of plum rain season, TMA(T late ,b) is the matrix of the predicted samples T(m,i,y) after time separation and classification, and the number of all samples with late plum rain end time is b i ; Get the final scene sum: a i +b j =λ; S32, using PMA method to analyze the matrix TMA (T early ,α) and the matrix TMA (T late ,b) Perform spatial stratification respectively, classify the samples according to the different height characteristics of the circulation field, and set the number of spatial stratification scenes to β, then the calculation formula is: Among them, PMA(TMA(T early ,α),e) is the predicted sample TMA (T early ,α) After the spatial stratification, the total number of spatial scenes is e: Among them, PMA(TMA(T late ,b) and f) are the predicted samples TMA (T late ,b) The matrix after spatial stratification, the total number of spatial scenes is f: The final scene sum is: e i +f j =b; S33. Use the methods in S31 and S32 to establish a multi-layer meteorological factor index matrix PMA (TMA,o).

5. The method for predicting the end of the plum rain season in the middle reaches of the Yangtze River based on a progressive strategy according to claim 4, characterized in that: The S4 is specifically: S41. According to the correlation rate of different climate factors, different weights ω are assigned to different climate factors: According to the different weights of different climate factors, calculate the weighted average final contribution rate of each key climate factor Set the trigger conditions of climate factors, When or When the rainy season ends, the corresponding climate scene will be automatically triggered; S42. Based on the meteorological factor indicator matrix PMA (TMA, e), meteorological factor triggering conditions are set under different scenarios. When the meteorological factors meet the requirements of a specific scenario, the corresponding end of the plum rain season date can be predicted.

Images