Quantitative assessment method and system for extreme rainfall attribution based on large-scale climate teleconnections

Abstract

The present invention provides a quantitative assessment method and system for extreme rainfall attribution based on large-scale climate teleconnection, the method comprising: step 1, dividing the study period into a reference period and a modern period; step 2, performing correlation analysis on the climate factors of 0 to 12 months before the extreme rainfall sequence and the extreme rainfall sequence, to obtain preliminarily screened climate factors; step 3, further screening the climate factors using stepwise regression, and incorporating them into the optimal explanatory variable set of the regression model; step 4, constructing a regression model describing the extreme rainfall sequence in the reference period based on the optimal explanatory variable set using multiple linear regression; step 5, using the regression model, introducing the climate factors of the modern period to reconstruct and simulate the extreme rainfall sequence; step 6, obtaining the frequency of the extreme rainfall event through frequency analysis, and then obtaining the recurrence period; step 7, comparing the simulated sequence and the observed sequence of the modern period to obtain the contribution rate of climate factors and human activities to a specific extreme rainfall event.

Description

Technical Field

[0001] The present invention belongs to the technical field of reservoir scheduling and operation, and in particular relates to a quantitative assessment method and system for extreme rainfall attribution based on large-scale climate teleconnection. Background Art

[0002] In recent decades, extreme precipitation events have been increasing globally. As the most active factor in the hydrological cycle, extreme changes in precipitation can severely impact the hydrological cycle, leading to a range of impacts on soil, vegetation, rivers, and human production and daily life. Furthermore, extreme precipitation events often trigger other extreme hydrological events, such as extreme floods and droughts, as well as major risk events such as urban waterlogging. Consequently, extreme precipitation has garnered increasing attention worldwide in recent years. Numerous studies have demonstrated a link between large-scale climate patterns and regional extreme hydrological phenomena. Furthermore, increasing urbanization and global warming are also significant factors contributing to the frequency of extreme rainfall events. Vast areas are susceptible to changes in sea temperature in the tropical Indian Ocean, which can trigger extreme rainfall. For these reasons, it is necessary to investigate the dominant factors driving extreme rainfall and quantify their impacts from multiple perspectives, including large-scale atmospheric circulation and air-sea interactions.

[0003] The following problems are common in many existing studies: (1) Few climate indicators are considered, and most of them select the most appropriate climate factor from several known climate factors that may have a significant impact on extreme rainfall in a certain area to explain extreme rainfall. This will lead to the neglect of other unknown significant climate factors and make the results slightly subjective; (2) Due to the few indicators considered, there are fewer explanatory variables in the prediction model, which cannot best describe the extreme rainfall sequence; (3) The analysis of the impact of climate control on extreme rainfall is based on the perspective of climate change in the modern period relative to the base period. Although this can quantify the extreme precipitation caused by climate change in the entire modern period, it is based on an entire period of time, and thus the contribution rate of a specific extreme rainfall cannot be obtained; therefore, existing research cannot accurately reveal the causes of extreme rainfall, and cannot achieve scientific research, prediction and effective prevention and control of extreme rainfall. Summary of the Invention

[0004] The present invention is made to solve the above problems, and its purpose is to provide a quantitative assessment method and system for extreme rainfall attribution based on large-scale climate teleconnections, which can fully consider the time lag of climate factors and specifically quantify each extreme rainfall event.

[0005] In order to achieve the above object, the present invention adopts the following scheme:

[0006] <Method>

[0007] like Figure 1As shown, the present invention provides a quantitative assessment method for extreme rainfall attribution based on large-scale climate teleconnection, which is characterized by comprising the following steps:

[0008] Step 1: Sample the longest and most continuous rainfall observation data in the study area to obtain the annual maximum value series for one day and the annual maximum value series for three consecutive days as the study series; use the non-parametric change point detection method to find the existing change points in the entire study period; based on the change points, define the period before the change point as the baseline period, and the period after the change point as the modern period;

[0009] Step 2: Use as many large-scale climate indicators as possible to conduct correlation analysis with extreme rainfall series during the reference period: It is assumed that the extreme rainfall obtained by sampling occurs in the months with concentrated precipitation in the study area. In order to fully consider the time lag effect of climate factors on extreme rainfall, such as Figure 2 As shown in the figure, the climate factors (C) from 0 to 12 months before the extreme rainfall sequence are respectively correlated with the extreme rainfall sequence. The climate factors entering the subsequent steps should meet the requirements of strong correlation, high significance level, continuous and sequence length that meets the research needs, and obtain multiple climate factors that are preliminarily screened;

[0010] Step 3. Generally speaking, if there are a large number of climate factors that have been obtained through step 2, and all of them are included in the model as explanatory variables, the model will become very complicated. When there are a large number of optional independent variables, some independent variables will inevitably have the following problems: (1) they are not significant relative to other independent variables; (2) they have multicollinearity with other independent variables; (3) they have large measurement errors. Therefore, based on the preliminary screening, stepwise regression is used to further screen out climate factors, and the climate factors that can ensure their significance, eliminate multicollinearity, and best describe the explained variables are obtained as the optimal explanatory variables, and are included in the optimal explanatory variable set of the regression model. Through steps 2 and 3, the best explanatory variable set for describing extreme rainfall is obtained from among many climate factors.

[0011] Step 4: Use multiple linear regression to construct a regression model describing the extreme rainfall series during the baseline period, using the climate factors obtained in step 3 as explanatory variables;

[0012] Step 5: Using the regression model, the modern climate factors are introduced to reconstruct and simulate the extreme rainfall sequence (modern simulation sequence);

[0013] Step 6: Obtain the frequency of the extreme rainfall event through frequency analysis, and then obtain the recurrence period;

[0014] Step 7. Based on the bootstrap method and Pearson III distribution, obtain the rainfall design values corresponding to the frequency of this extreme rainfall event for the modern simulation series and observation series. Compare the design values of the modern simulation series and observation series to obtain the contribution rate of climate factors and human activities to a specific extreme rainfall event.

[0015] Preferably, the quantitative assessment method for attribution of extreme rainfall based on large-scale climate teleconnections provided by the present invention may also have the following characteristics: in step 1, at least two non-parametric change point detection methods are used to search for existing change points throughout the entire study period; if the results of the various non-parametric change point detection methods all meet their respective significance requirements, the common result is selected as the mutation point; if there is no common result, a point that meets the significance requirements is selected based on the actual situation of the study period (for example, urbanization data of the years before and after the change point to be determined can be referenced: such as urban area growth, the proportion of tertiary industry added value in GDP, etc.); if there is a case where significance is not met, a change point that meets the significance requirements should be selected based on the actual situation. In addition, the results of identifying the mutation point for the annual maximum value and the annual maximum value for three consecutive days are likely to be different, and it is necessary to discuss the annual maximum value and the annual maximum value for three consecutive days separately according to different change points.

[0016] For example, the Mann–Kendall and Pettitt mutation point tests are used to verify each other to determine the existence of the change point. For the Mann–Kendall mutation point test, the significance level is α = 0.05, and the critical value of the statistic is ±1.96. The Pettitt mutation point test must meet P ≤ 0.5:

[0017]

[0018] In the formula And s k is the statistic obtained by Pettitt's mutation point test.

[0019] Preferably, the quantitative assessment method of extreme rainfall attribution based on large-scale climate teleconnection provided by the present invention may also have the following characteristics: in step 2, the selected climate factors should satisfy the Pearson correlation coefficient | r xy |≥0.3, significance level p≤0.05, the series is long enough to fit the long series precipitation data and there are not many missing observations.

[0020] Preferably, the quantitative assessment method for attribution of extreme rainfall based on large-scale climate teleconnections provided by the present invention may also have the following characteristics: in step 6, the difference between the modern observation data and the simulated data is taken as the impact of human activities; the contribution rate of climate factors and human activities to extreme rainfall is obtained by the percentage of their contribution values to the sum of their absolute values:

[0021] p t =|p c |+|p h |,

[0022]

[0023]

[0024] Where r c and r h represent the contribution rates of climate factors and human activities to extreme rainfall; p c and p h represent the contribution of climate factors and human activities to extreme rainfall; p t It represents the sum of the absolute values of the two contributions.

[0025] Preferably, the quantitative assessment method for attribution of extreme rainfall based on large-scale climate teleconnection provided by the present invention may also have the following characteristics: in step 7, the Bootstrap method is used to sample the entire study period, the baseline period, and the modern period N times, N ≥ 10000, and the N groups of new sequences in each case are fitted using the Pearson III curve, mainly obtaining rainfall with probabilities of 10%, 1%, 0.1%, and 0.01%, corresponding to extreme rainfall with a return period of once in ten years, once in a hundred years, once in a thousand years, and once in ten thousand years, and making interval estimates; the contribution rate is calculated by the design value of extreme rainfall under the corresponding probability obtained by the bootstrap and Pearson III distribution, so that the result is more accurate and reasonable.

[0026] <system>

[0027] Furthermore, the present invention also provides a quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnections, which is characterized by comprising:

[0028] The division section samples the longest and most continuous rainfall observation data in the study area to obtain the annual one-day maximum value series and the annual three-day maximum value series as the study series; uses the non-parametric change point detection method to find the existing change points in the entire study period; based on the change points, the period before the change point in the study period is defined as the baseline period, and the period after the change point in the study period is defined as the modern period;

[0029] In the preliminary screening phase, correlation analysis is performed between as many large-scale climate indicators as possible and the extreme rainfall series during the baseline period. To fully consider the time lag effect of climate factors on extreme rainfall, correlation analysis is performed between the climate factors from 0 to 12 months before the extreme rainfall series. The climate factors that enter the subsequent steps should meet the requirements of strong correlation, high significance level, and continuous sequence length that meets the research requirements, thus obtaining the preliminary screened climate factors.

[0030] In the optimal factor screening part, stepwise regression is used to further screen out climate factors, and the climate factors that can both ensure their significance and best describe the explained variables are obtained as the optimal explanatory variables and included in the optimal explanatory variable set of the regression model;

[0031] The model building department uses multiple linear regression to construct a regression model describing the extreme rainfall series in the reference period with the climate factors obtained by the optimal factor screening department as explanatory variables;

[0032] The reconstruction department uses regression models to bring in modern climate factors to reconstruct and simulate extreme rainfall sequences;

[0033] The recurrence period acquisition unit obtains the frequency of the extreme rainfall event through frequency analysis, and then obtains the recurrence period;

[0034] The contribution rate calculation part uses the bootstrap method and Pearson III distribution to obtain the rainfall design values corresponding to the frequency of this extreme rainfall event for the modern simulation series and observation series. By comparing the design values of the modern simulation series and observation series, the contribution rate of climate factors and human activities to a specific extreme rainfall event is obtained.

[0035] The control unit is connected to the division unit, preliminary screening unit, optimal factor screening unit, model building unit, reconstruction unit, contribution rate calculation unit, and recurrence period acquisition unit to control their operations.

[0036] Preferably, the quantitative assessment system for attribution of extreme rainfall based on large-scale climate teleconnection provided by the present invention may also include: an input display unit, which is communicatively connected to the division unit, preliminary screening unit, optimal factor screening unit, model building unit, reconstruction unit, contribution rate calculation unit, recurrence period acquisition unit, and control unit, and is used to allow the operator to input control instructions and perform corresponding displays.

[0037] Preferably, the quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnection provided by the present invention may also include: an early warning unit, which is communicatively connected to both the control unit and the input and display unit, and generates corresponding early warning graphic information or issues corresponding early warning signals to provide danger warnings based on the results of the contribution rate calculation unit and the recurrence period acquisition unit for risks that may be caused by the recurrence period.

[0038] Preferably, the quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnection provided by the present invention may also include: an extreme rainfall prediction unit, which obtains forecast information of future rainfall and climate factors, and uses this information as future rainfall observation data and climate factors to calculate the future rainfall conditions and contribution rates caused by future climate factors and human activities using the division unit, the preliminary screening unit, the optimal factor screening unit, the model building unit, the reconstruction unit, the recurrence period acquisition unit, and the contribution rate calculation unit, and predicts future extreme rainfall events and causes based on this.

[0039] Preferably, the quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnection provided by the present invention may also have the following characteristics: in the division part, at least two non-parametric change point detection methods are used to find the existing change points in the entire study period. If the results of various non-parametric change point detection methods meet their respective significance requirements, the common result is selected as the mutation point; when there is no common result, the point that meets the significance requirement is selected according to the actual situation of the study period; if there is a situation that does not meet the significance, it should be selected according to the actual situation among the change points that meet the significance requirement.

[0040] Preferably, the quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnection provided by the present invention may also have the following characteristics: in the preliminary screening part, the selected climate factors should satisfy the Pearson correlation coefficient | r xy |≥0.3, significance level p≤0.05, the series is long enough to fit the long series precipitation data and there are not many missing observations.

[0041] Preferably, the quantitative assessment system for attribution of extreme rainfall based on large-scale climate teleconnections provided by the present invention may also have the following features: in the contribution rate calculation unit, the difference between modern observation data and simulation data is used as the impact of human activities; the contribution rates of climate factors and human activities to extreme rainfall are obtained by the percentage of their contribution values to the sum of their absolute values:

[0042] p t =|p c |+|p h |,

[0043]

[0044]

[0045] Where r c and r h represent the contribution rates of climate factors and human activities to extreme rainfall; p c and p h represent the contribution of climate factors and human activities to extreme rainfall; pt It represents the sum of the absolute values of the two contributions.

[0046] Functions and effects of the invention

[0047] The quantitative assessment method and system for extreme rainfall attribution based on large-scale climate teleconnections provided by the present invention use the longest possible measured rainfall data to sample and derive an extreme rainfall sequence. The occurrence time of the extreme rainfall sequence is processed according to the precipitation concentrated months in the study area, and the study sequence is divided into a baseline period sequence affected only by climate factors and a modern period sequence affected by both climate factors and human activities. The influence of as many climate indicators as possible on extreme rainfall is considered and optimized through statistical methods. The most significant and statistically significant climate factors are obtained using multiple climate indicators after two-step optimization. A model is derived through multivariate linear regression to construct a modern period simulation sequence. The simulated rainfall is predicted or estimated by the optimal combination of multiple climate indicators. The various climate indicators are utilized to the maximum extent to derive predictions or estimates of extreme rainfall under climate change. Instead of quantitative attribution from the perspective of climate change, the system specifically quantifies a particular extreme rainfall event while fully considering the time lag of climate factors. This ultimately improves the level of explanation of extreme rainfall by climate factors, reduces subjectivity, and makes the results more objective and accurate.

[0048] In addition, the present invention first calculates the sum of the absolute values of the contributions of the two to a certain extreme rainfall, and then quickly quantifies the contribution rate in the form of the percentage of the contribution value to the sum of the absolute values. This can not only reliably quantify the attribution, but also improve the calculation and processing efficiency of the present invention.

[0049] In summary, the quantitative assessment method and system for extreme rainfall attribution based on large-scale climate teleconnections provided by the present invention have constructed a method system for studying the dominant attribution and contribution of climate control and human activities to a specific extreme rainfall event. It can simply and effectively analyze which of climate factors and human activities is the dominant factor in a specific extreme rainfall event, and takes into account the actual impact of climate indicators and their time lags on extreme rainfall at the largest scale, which is conducive to scientific and effective research, risk warning and prevention and control of extreme rainfall. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flowchart of a quantitative assessment method for extreme rainfall attribution based on large-scale climate teleconnection according to the present invention;

[0051] Figure 2 This is a schematic diagram of the present invention taking into account the time lag of climate factors;

[0052] Figure 3 Schematic diagram of the mutation point test of the 71-year extreme rainfall series at Zhengzhou Station involved in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The specific implementation scheme of the quantitative assessment method and system for extreme rainfall attribution based on large-scale climate teleconnection according to the present invention is described in detail below with reference to the accompanying drawings.

[0054] <Example>

[0055] like Figure 1 As shown, the quantitative assessment method for extreme rainfall attribution based on large-scale climate teleconnection provided in this embodiment includes the following steps:

[0056] Step 1. Obtain the longest and most continuous rainfall sequence in the study area and sample it to obtain the annual maximum value sequence for one day and the annual maximum value sequence for three consecutive days as the research data. The change point is determined by the Mann–Kendall mutation point test and the Pettitt mutation point test. For the Mann–Kendall mutation point test:

[0057]

[0058]

[0059]

[0060] Where n is the sample size; x is the extreme rainfall sequence; E(s k ) and var(s k ) are the statistics s k The mean and variance of :

[0061]

[0062]

[0063] Reverse the sequence x, repeat the above calculation and define the result as:

[0064] UB k =-UF k , k=n,n-1,...,1

[0065] Take the significance level a=0.05, then the critical value is ±1.96, and UF k UB k and the critical value representing the significance level are plotted together in a graph. If the UF k and UB k If they intersect, then the intersection point can be considered as the change point during the study period. For the Pettitt mutation point test, r i The discussion is divided into three situations:

[0066]

[0067]

[0068] In the formula, if the statistic s at time t0 k If the absolute value reaches its maximum value and its corresponding P≤0.5, it is considered that there is a mutation point within the change period at time t0:

[0069]

[0070] After obtaining the mutation point within the study period, the period before the change point is defined as the "baseline period" and the period after the change point is defined as the "modern period".

[0071] Step 2. During the baseline period, correlation analysis is performed between the baseline extreme rainfall series and as many climate indicators as possible (including hundreds of climate factors, such as the North Atlantic Oscillation Index, El Niño Temperature Index, Sunspot Index, etc.):

[0072]

[0073] Where r xy is the Pearson correlation coefficient between two variables, when |r xy When |<0.1, it is considered that there is no correlation between the two variables. xy When |<0.3, it is considered that the two variables have a weak correlation. xy When |<0.5, it is considered that there is a medium correlation between the two variables. xy When |≥0.5, the correlation between the two variables is considered to be very strong; x and σ y are the sample standard deviations of x and y respectively; Cov(x,y) represents the sample covariance:

[0074]

[0075] We selected climate factors with moderate or higher correlations and verified their significance. Only factors that met the p < 0.05 criteria were selected. Factors that met the correlation and significance requirements should be discarded if they had a large number of missing data.

[0076] Step 3. Further optimize climate indicators through stepwise regression to obtain the best explanatory variable set describing the extreme rainfall series:

[0077] (1) First, establish simple linear regression for all explanatory variables and explained variables and perform F test, find the explanatory variable with the largest F value and introduce it into the model;

[0078] (2) Introduce other explanatory variables one by one. Calculate the partial regression sum of squares of all independent variables outside the model and select the maximum value to perform a significance test at a given level. If the significance requirement is met, then introduce the independent variable into the model;

[0079] (3) When a new independent variable is introduced into the model, it is necessary to recalculate the partial regression sum of squares of all independent variables in the model and select the minimum value to perform a significance test at a given level. If significant, all independent variables in the model are retained. If not significant, the independent variable is removed and tested again. Only when all independent variables in the model meet the significance requirements can the next new independent variable be introduced.

[0080] (4) Continue (3) and (4) until no variables can be eliminated from the model and no new variables can be introduced, and the final set of explanatory variables is obtained;

[0081] Among them, the partial regression sum of squares refers to the part of the regression sum of squares of the entire model that is reduced after removing an independent variable in the multiple regression model:

[0082]

[0083] Where, ESS is the regression sum of squares; is the regression value of the dependent variable; is the mean value of the dependent variable.

[0084] Step 4. Use the selected optimal explanatory variable set to construct a multiple linear regression model to explain extreme rainfall:

[0085] y i =β0+β1x i1 +β2x i2 +…+β m x im +e i ,

[0086] E(y)=β0+β1x1+β2x2+…+β m x m ,

[0087] Where E(y) is the theoretical regression equation of multiple linear regression.

[0088] Step 5. Use the regression model to reconstruct a simulated series of extreme rainfall in the modern period. While extreme rainfall in the baseline period was influenced solely by climate factors, extreme rainfall in the modern period results from the combined effects of climate factors and human activities. Therefore, the simulated series reflects the impact of modern climate factors on extreme rainfall.

[0089] Step 6. Use frequency analysis to determine the frequency of a specific extreme rainfall event and then the recurrence period. Choose the Pearson III distribution, which is widely used in hydrological frequency analysis in China:

[0090]

[0091] Where Γ(α) represents the gamma function of α; α is the shape parameter; β is the scale parameter; a0 is the location parameter; α>0, β>0.

[0092] For heavy rain events, the return period is defined as:

[0093]

[0094] Where T represents the recurrence period in years.

[0095] Step 7. Bootstrap the observed and simulated series for the modern period 10,000 times, generating 10,000 new series. Fit the new series using a Pearson III distribution to obtain design values and interval estimates for rainfall at different frequencies. Compare the design value differences between the simulated and observed series for the modern period to determine the contribution of climate change and human activities to a specific extreme rainfall event.

[0096] The modern extreme rainfall sequence is composed of two parts: rainfall controlled by climate and rainfall controlled by human activities.

[0097] p=p c +p h ,

[0098] Where p represents the extreme rainfall observation value; p h It represents the contribution of human activities (e.g. urban area, number of large-scale water conservancy projects, population, farmland area, etc.) to extreme rainfall; c Indicates extreme rainfall values under climate control.

[0099] The contribution rate can be calculated as follows:

[0100] p t =|p c |+|p h |,

[0101]

[0102]

[0103] Where r c and r hRespectively represent the contribution rates of climate factors and human activities to extreme rainfall.

[0104] In this embodiment, the heavy rainstorm in Zhengzhou on July 20 is taken as an example to compare the method of the present invention with the algorithm of the prior art.

[0105] The method of the present invention is used to sample the historical long series of daily rainfall data of Zhengzhou Station from 1951 to 2021, and the annual maximum rainfall sequence and the annual average maximum rainfall sequence for three consecutive days are obtained, such as Figure 3 As shown, using two breakpoint detection methods, the breakpoint for the annual maximum daily rainfall series occurred in 2004, while the breakpoint for the annual three-day average maximum rainfall series occurred in 2002. Based on the breakpoints, the entire study period was divided into a baseline period and a modern period. During the baseline period, 130 climate factors were correlated with the extreme rainfall series and initially selected according to step 2. Then, a set of explanatory variables was constructed according to step 3. Frequency analysis determined the frequency of this extreme rainfall event to be 1%. According to step 7, the bootstrap method was used to replicate the annual maximum daily rainfall series and the annual three-day average maximum rainfall series 10,000 times in each of the baseline, modern, and entire study periods, generating 10,000 new extreme rainfall series in each case. Each new series was fitted using a Pearson III distribution to obtain design rainfall values and interval estimates corresponding to 10%, 1%, 0.1%, and 0.01%. As shown in Table 1, the calculated design value of the annual maximum one-day rainfall corresponding to a 1% frequency under climate control is 141.40±45.97 mm, and the design value of the annual maximum three-day rainfall is 46.65±7.93 mm; the contribution of climate factors to this extreme rainfall is 79.79% and 60.85%, respectively.

[0106] Using existing methods, a linear simulation would yield extreme rainfall amounts of 61.99 mm and 20.15 mm, respectively. Human activity would contribute as much as 93.41% and 94.97% to these rainfall events, a clearly unreasonable result. This is due to the fact that, in the absence of significant mutations in climate factors, the shortfall in the simulated values relative to the observed values is attributed to human activity. Clearly, the attribution analysis of extreme rainfall using existing methods, compared to the present invention, suffers from significant uncertainty and irrationality. Such results, based on existing methods, are insufficient for scientific and rational research, risk warning, and prevention and control of extreme rainfall.

[0107] Table 1 Comparison of the contribution of human activities to the extreme rainfall in Zhengzhou on July 20 between the present invention and the prior art methods

[0108]

[0109] Furthermore, the above method can be automatically controlled and implemented through a quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnection, which includes a division part, a preliminary screening part, an optimal factor screening part, a model building part, a reconstruction part, a recurrence period acquisition part, a contribution rate calculation part, an early warning part, an input display part, and a control part.

[0110] The division section performs the contents described in step 1 above, sampling the longest and most continuous rainfall observation data in the study area to obtain the annual one-day maximum value series and the annual three-day maximum value series as the study series; uses the non-parametric change point detection method to find the existing change points in the entire study period; based on the change points, the period before the change point in the study period is defined as the baseline period, and the period after the change point in the study period is defined as the modern period.

[0111] The preliminary screening part performs the contents described in step 2 above, and uses as many large-scale climate indicators as possible to conduct correlation analysis with the extreme rainfall sequence during the reference period: In order to fully consider the time lag effect of climate factors on extreme rainfall, the climate factors of 0 to 12 months before the extreme rainfall sequence are respectively subjected to correlation analysis with the extreme rainfall sequence. The climate factors entering the subsequent steps should meet the requirements of strong correlation, high significance level, continuity and sequence length that meets the research requirements, so as to obtain the preliminary screened climate factors.

[0112] The optimal factor screening unit performs the steps described in step 3 above and uses stepwise regression to further screen out climate factors. The climate factors that can both ensure their significance and best describe extreme rainfall are obtained as the optimal explanatory variables and are included in the optimal explanatory variable set of the regression model.

[0113] The model building section performs the steps described in step 4 above, and uses multiple linear regression to construct a regression model describing the extreme rainfall series in the baseline period with the climate factors obtained by the optimal factor screening section as explanatory variables.

[0114] The reconstruction part uses the regression model to perform the content described in step 5 above, bringing in modern climate factors to reconstruct and simulate the extreme rainfall sequence.

[0115] The recurrence period acquisition unit executes the contents described in step 6 above, obtains the frequency of the extreme rainfall event through frequency analysis, and then obtains the recurrence period.

[0116] The contribution rate calculation unit performs the steps described in step 7 above, comparing the simulated and observed series in the modern period to derive the contribution rates of climate change and human activities to a specific extreme rainfall event.

[0117] The extreme rainfall prediction unit obtains forecast information of future rainfall and climate factors, uses this information as future rainfall observation data and climate factors, and uses the division unit, the preliminary screening unit, the optimal factor screening unit, the model construction unit, the reconstruction unit, the recurrence period acquisition unit, and the contribution rate calculation unit to calculate the future rainfall conditions and contribution rates caused by future climate factors and human activities, and predicts future extreme rainfall events and causes based on this.

[0118] Based on the results of the contribution rate calculation unit, the recurrence period acquisition unit and the extreme rainfall prediction unit, the early warning unit generates corresponding early warning graphic information or issues corresponding early warning signals to provide danger warnings for possible dangers in the future; further, the early warning unit can also display or prompt operators to determine the safety factor or level that buildings or projects need to achieve based on future extreme rainfall events or possible dangers in the future.

[0119] The input display unit is used to allow the operator to input control instructions and display them accordingly.

[0120] The control unit is communicatively connected with the division unit, preliminary screening unit, optimal factor screening unit, model building unit, reconstruction unit, contribution rate calculation unit, recurrence period acquisition unit, extreme rainfall prediction unit, early warning unit, and input display unit to control their operations.

[0121] The above embodiments are merely illustrative of the technical solutions of the present invention. The quantitative assessment method and system for extreme rainfall attribution based on large-scale climate teleconnections involved in the present invention are not limited solely to the contents described in the above embodiments, but are subject to the scope defined by the claims. Any modifications, supplements, or equivalent substitutions made by persons skilled in the art based on these embodiments are within the scope of protection claimed by the claims.

Claims

1. A quantitative assessment method for attribution of extreme rainfall based on large-scale climate teleconnections, characterized by: The following steps are involved: Step 1: Sample the longest and most continuous rainfall observation data in the study area to obtain the annual maximum value series and the annual maximum value series of three consecutive days as the study series; use the non-parametric change point detection method to find the existing change points in the entire study period; Based on the change point, the period before the change point is defined as the baseline period, and the period after the change point is defined as the modern period. Step 2: Conduct correlation analysis between as many large-scale climate indicators as possible and the extreme rainfall series during the baseline period. To fully account for the time lag effect of climate factors on extreme rainfall, conduct correlation analysis between the climate factors from 0 to 12 months before the extreme rainfall series and the extreme rainfall series. Climate factors to be selected for subsequent steps should meet the requirements of strong correlation, high significance level, and continuous sequence length that meets the research requirements, thus obtaining preliminary screening of climate factors. Step 3: Use stepwise regression to further screen out climate factors, and obtain the climate factors that can both ensure their significance and best describe extreme rainfall as the optimal explanatory variables, and include them in the optimal explanatory variable set of the regression model; Step 4: Use multiple linear regression to construct a regression model describing the extreme rainfall series during the baseline period, using the climate factors obtained in step 3 as explanatory variables; Step 5: Use the regression model to bring in modern climate factors to reconstruct and simulate the extreme rainfall sequence; Step 6: Obtain the frequency of the extreme rainfall event through frequency analysis, and then obtain the recurrence period; Step 7. Based on the bootstrap method and Pearson III distribution, obtain the rainfall design values corresponding to the frequency of this extreme rainfall event for the modern simulation series and observation series. Compare the design values of the modern simulation series and observation series to obtain the contribution rate of climate factors and human activities to a specific extreme rainfall event.

2. The quantitative assessment method for extreme rainfall attribution based on large-scale climate teleconnections according to claim 1 is characterized by: in, In step 1, at least two non-parametric change point detection methods are used to search for existing change points in the entire study period. If the results of various non-parametric change point detection methods all meet their respective significance requirements, the common result is selected as the mutation point; if there is no common result, the point that meets the significance requirements is selected according to the actual situation of the study period; if there is a situation that does not meet the significance, it should be selected according to the actual situation among the change points that meet the significance requirements.

3. The quantitative assessment method for extreme rainfall attribution based on large-scale climate teleconnections according to claim 1 is characterized by: in, In step 2, the climate factors selected should satisfy the Pearson correlation coefficient , significance level , the sequence is long enough to be compatible with the long series of precipitation data and there are not many missing measurements.

4. The quantitative assessment method for extreme rainfall attribution based on large-scale climate teleconnections according to claim 1 is characterized by: in, In step 7, the difference between the modern observation data and the simulated data is taken as the impact of human activities; the contribution rate of climate factors and human activities to extreme rainfall is obtained by the percentage of their contribution value to the sum of absolute values: ， ， ， Where, and Respectively represent the contribution rates of climate factors and human activities to extreme rainfall; and Respectively represent the contribution of climate factors and human activities to extreme rainfall; It represents the sum of the absolute values of the two contributions.

5. The quantitative assessment method for extreme rainfall attribution based on large-scale climate teleconnections according to claim 1 is characterized by: in, In step 7, the Bootstrap method is used to sample the entire study period, baseline period, and modern period N times, where N ≥ 10,000. The N new series in each case are fitted using the Pearson III curve to obtain rainfall with probabilities of 10%, 1%, 0.1%, and 0.01%, corresponding to extreme rainfall events of once in a decade, once in a century, once in a thousand years, and once in ten thousand years, and interval estimates are made; the contribution rate is calculated using the design value of extreme rainfall under the corresponding probabilities obtained from the bootstrap and Pearson III distributions.

6. A quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnections, characterized by: include: The division department samples the longest and most continuous rainfall observation data in the study area to obtain the annual one-day maximum value sequence and the annual three-day maximum value sequence as the study sequence; Nonparametric change point detection methods were used to find the existing change points throughout the study period; Based on the change point, the period before the change point is defined as the baseline period, and the period after the change point is defined as the modern period. In the preliminary screening phase, correlation analysis is performed between as many large-scale climate indicators as possible and the extreme rainfall series during the baseline period. To fully consider the time lag effect of climate factors on extreme rainfall, correlation analysis is performed between the climate factors from 0 to 12 months before the extreme rainfall series and the extreme rainfall series. The climate factors that enter the subsequent steps should meet the requirements of strong correlation, high significance level, and continuous sequence length that meets the research requirements, thus obtaining the preliminary screening climate factors. In the optimal factor screening part, stepwise regression is used to further screen out climate factors, and the climate factors that can both ensure their significance and best describe extreme rainfall are obtained as the optimal explanatory variables, which are then included in the optimal explanatory variable set of the regression model; The model building unit uses multiple linear regression to build a regression model describing the extreme rainfall series in the reference period using the climate factors obtained by the optimal factor screening unit as explanatory variables; The reconstruction department uses regression models to bring in modern climate factors to reconstruct and simulate extreme rainfall sequences; The recurrence period acquisition unit obtains the frequency of the extreme rainfall event through frequency analysis, and then obtains the recurrence period; The contribution rate calculation part uses the bootstrap method and Pearson III distribution to obtain the rainfall design values corresponding to the frequency of this extreme rainfall event for the modern simulation series and observation series. By comparing the design values of the modern simulation series and observation series, the contribution rate of climate factors and human activities to a specific extreme rainfall event is obtained. The control unit is connected to the division unit, the preliminary screening unit, the optimal factor screening unit, the model construction unit, the reconstruction unit, the recurrence period acquisition unit, and the contribution rate calculation unit to control their operations.

7. The quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnection according to claim 6, characterized in that: Also includes: The input display unit is communicatively connected with the division unit, the preliminary screening unit, the optimal factor screening unit, the model building unit, the reconstruction unit, the recurrence period acquisition unit, the contribution rate calculation unit, and the control unit, and is used to allow the operator to input control instructions and perform corresponding displays.

8. The quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnection according to claim 7, characterized in that: Also includes: The early warning unit is communicatively connected with the control unit and the input and display unit, and generates corresponding early warning graphic information or issues corresponding early warning signals to provide danger warnings based on the results of the contribution rate calculation unit and the recurrence period acquisition unit, in response to the dangers that may be caused by the recurrence period.

9. The quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnection according to claim 6, characterized in that: Also includes: The extreme rainfall prediction unit obtains forecast information of future rainfall and climate factors, and uses this information as future rainfall observation data and climate factors to calculate the future rainfall conditions and contribution rates caused by future climate factors and human activities using the division unit, the preliminary screening unit, the optimal factor screening unit, the model construction unit, the reconstruction unit, the recurrence period acquisition unit, and the contribution rate calculation unit, and predicts future extreme rainfall events based on this.

10. The quantitative assessment system for extreme rainfall attribution based on large-scale climate teleconnections according to claim 6, characterized in that: in, In the division section, at least two non-parametric change point detection methods are used to search for existing change points in the entire study period. If the results of various non-parametric change point detection methods meet their respective significance requirements, the common result is selected as the mutation point; when there is no common result, the point that meets the significance requirement is selected according to the actual situation of the study period; if there is a situation that does not meet the significance, it should be selected according to the actual situation among the change points that meet the significance requirement.

Images