A method for predicting waterlogging on sunken bridges based on downscaling and dynamic division of catchment areas

Abstract

The present invention relates to the field of urban waterlogging control, and in particular to a method for predicting water accumulation in a sunken bridge based on downscaling and dynamic division of catchment areas. S10: Predicting rainfall in the sunken bridge area based on large-scale climate forecast factors in the sunken bridge area; S20: Dividing the sunken bridge area into a main catchment area and multiple secondary catchments, and obtaining elevation data for the main catchment area and each secondary catchment area; S30: Determining the risk rating of each secondary catchment area based on rainfall in the sunken bridge area, the elevation of the main catchment area, and the elevation data of each secondary catchment area; S40: Obtaining the amount of water flowing into the main catchment area from each secondary catchment area based on the risk rating of each secondary catchment area; S50: Obtaining the total amount of water accumulated in the main catchment area based on the amount of water flowing into the main catchment area from each secondary catchment area and the amount of water accumulated in the main catchment area itself. The present invention can accurately predict rainfall in the sunken bridge area, and at the same time, dividing the catchment area according to the risk rating can reduce computer computing load and improve prediction accuracy.

Description

Technical Field

[0001] The present invention relates to the field of urban waterlogging control, and in particular to a method for predicting waterlogging on a sunken bridge based on downscaling and dynamic division of a catchment area. Background Art

[0002] The large number of sunken bridges constructed during my country's urbanization has become a frequent and vulnerable area for flooding during flood season, impacting regional transportation and public safety. With changes in the natural environment, extreme weather events are becoming more frequent, seriously impacting public safety. When urban flooding occurs, the areas surrounding sunken bridges are low-lying, making them prone to deep ponding if drainage is poor. Furthermore, since sunken bridges are often located along main arterial roads, the resulting damage is significant, impacting the lives and work of urban residents and even causing casualties. Cities have a high concentration of hardened surfaces, especially around main arterial roads, leaving little room for water to collect. Sunken bridges are located at low elevations, but the actual water storage capacity within these areas is limited. When flooding occurs, water accumulates rapidly, reaching dangerous levels in just a few minutes.

[0003] Currently, flooding predictions for sunken bridge areas are typically determined by linking all catchment areas with a prediction model. This linking of all catchment areas increases the computer's computational load during periods of low rainfall, wasting computing power. Furthermore, existing prediction models cannot be flexibly applied to actual rainfall in the sunken bridge area. Summary of the Invention

[0004] The present invention provides a method for predicting water accumulation in a sunken bridge based on downscaling and dynamic division of a catchment area, which solves at least one technical problem existing in the prior art.

[0005] The technical solution adopted by the present invention is as follows: a method for predicting waterlogging on sunken bridges based on downscaling and dynamic division of catchment areas, comprising:

[0006] S10: Predict the rainfall in the area of ​​the sunken bridge;

[0007] S20: Divide the sunken bridge area into a main catchment area and multiple secondary catchments, and obtain elevation data of the main catchment area and each secondary catchment area;

[0008] S30: Determine the risk rating of each sub-catchment based on the rainfall in the Sag Bridge area, the elevation of the main catchment area, and the elevation data of each sub-catchment area;

[0009] S40: according to the risk rating of each sub-catchment area, obtain the water volume of each sub-catchment area into the main catchment area;

[0010] S50: Obtain the total accumulated water volume of the main catchment area according to the amount of water collected by each secondary catchment area into the main catchment area and the accumulated water volume of the main catchment area itself.

[0011] Furthermore, the step S10 includes:

[0012] S110: Establishing a downscaling model by using a statistical downscaling method, wherein the downscaling model is a statistical relationship between large-scale climate prediction factors and rainfall in the Xiasao Bridge area;

[0013] S120: Predict rainfall in the Xiaaoqiao area based on the downscaling model and large-scale climate prediction factors.

[0014] Furthermore, the step S110 includes:

[0015] S111: Obtain large-scale climate prediction factors and rainfall in the Xiaaoqiao area observed over many years, and establish a downscaling model using dynamic or statistical methods;

[0016] S112: Validate the downscaling model.

[0017] Furthermore, the step S112 includes:

[0018] Select the predictor variables suitable for analyzing the precipitation in Xiasaoqiao from all the large-scale climate predictors;

[0019] According to the actual data of the predictor variables observed during the calibration period and the downscaling model, a multivariate linear regression equation system is established;

[0020] The rainfall simulation is obtained based on the actual data of the predictor variables observed during the validation period and the multivariate linear regression equation system;

[0021] The downscaling model was tested using statistical correlation coefficients based on the observed rainfall and simulated rainfall during the validation period.

[0022] Furthermore, the statistical correlation coefficient includes a correlation coefficient and a Nash coefficient.

[0023] Furthermore, the large-scale climate prediction factors of the Xiaao Bridge area are obtained by weather monitoring stations built near the Xiaao Bridge.

[0024] Furthermore, the step S30 includes:

[0025] When the elevation data of the main catchment area is greater than that of the secondary catchment area, the formula is:

[0026]

[0027] Determine the risk factor of the catchment area, where K is the risk factor, R is the rainfall in the concave bridge area, and H is the 主 The elevation data of the main catchment area, H 次 It is the elevation data of the secondary catchment area;

[0028] Determine the risk rating based on the risk coefficient and the preset correspondence;

[0029] When the elevation data of the main catchment area is lower than that of the secondary catchment area, the risk rating of the secondary catchment area is level one.

[0030] Furthermore, the preset corresponding relationship includes:

[0031] When the risk factor is greater than the first preset value, the risk rating is level one;

[0032] When the risk factor is less than or equal to the first preset value and greater than the second preset value, the risk rating is the second level;

[0033] When the risk factor is less than or equal to the second preset value and greater than the third preset value, the risk rating is level three;

[0034] When the risk coefficient is less than or equal to the third preset value, the risk rating is level four.

[0035] Furthermore, the step S40 includes:

[0036] When the risk rating is level 1, the amount of water flowing from the secondary catchment into the main catchment is equal to the runoff generated by the underlying surface of the secondary catchment;

[0037] When the risk rating is level 2, the water volume of the secondary catchment area flowing into the main catchment area is equal to the runoff volume generated by the underlying surface of the secondary catchment area multiplied by the first preset multiple;

[0038] When the risk rating is level three, the amount of water flowing from the secondary catchment area into the main catchment area is equal to the runoff generated by the underlying surface of the secondary catchment area at a second preset multiple;

[0039] When the risk rating is level 4, the secondary catchment area does not feed water into the main catchment area.

[0040] Furthermore, the runoff generated by the underlying surface of the secondary catchment area and the water accumulation in the primary catchment area itself are both obtained through the urban sunken bridge waterlogging model.

[0041] Beneficial effects of the present invention: The present invention predicts the future rainfall in the sunken bridge area through the large-scale climate forecast factors of the sunken bridge area, divides the sunken bridge catchment area into regions, determines the risk rating of each return flow area, determines the water volume of the catchment area according to different ratings, and thus determines the water accumulation volume of the main catchment area. The present invention can accurately predict the rainfall in the sunken bridge area, and at the same time, the division according to the risk rating of different catchment areas can reduce the computer calculation load and improve the prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flow chart of the present invention.

[0043] Figure 2 It is a flow chart of rainfall prediction in the present invention.

[0044] Figure 3 It is a flow chart of the watershed risk rating in the present invention. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0046] In an embodiment of the present invention, Figure 1 This is a flowchart of the specific steps of a method for predicting waterlogging at sunken bridges based on downscaling and dynamic division of catchment areas according to the present invention. Figure 3 It is a flow chart of the watershed risk rating in the present invention.

[0047] like Figure 1 and Figure 3 As shown, the present invention includes:

[0048] S10: Predict the rainfall in the Sag Bridge area.

[0049] S110: A downscaling model is established through the statistical downscaling method. The location of the sunken bridge area is generally at the intersection of urban main roads. The rainfall in the sunken bridge area is generally predicted based on the atmospheric circulation model (GCM) and large-scale climate forecast factors. This prediction method is a conventional technique used by those skilled in the art, so it will not be repeated here.

[0050] In one embodiment of the present invention, a downscaling model can be used to accurately predict rainfall in the concave bridge area. The steps for accurately predicting rainfall in the concave bridge area are as follows:

[0051] The downscaling model is the statistical relationship between large-scale climate forecast factors and rainfall in the Xiaaoqiao area.

[0052] To obtain large-scale climate forecasts for the sunken bridge area, a weather monitoring station is typically set up near the area. These stations are constructed of corrosion-resistant materials, such as high-yield-strength carbon steel or stainless steel brackets, to ensure long-term operation in outdoor environments. Different meteorological sensors are selected based on site conditions. Data is transmitted via a local area network (LAN) or wireless network. LANs can be configured using modems, fiber optic networks, routers, and other components. Wireless networks can be categorized into three types: short-range, medium-range, and long-range transmission, depending on the communication distance. GPRS / 4G or GSM are generally used.

[0053] Large-scale climate forecast factors include air humidity, air temperature, atmospheric pressure, wind direction, wind speed, rainfall, evaporation, light and ultraviolet intensity, etc.

[0054] S120: Predict the rainfall in the Xiasao Bridge area based on the downscaling model.

[0055] Precipitation forecasts for the Xiaaoqiao region rely primarily on downscaling models. Statistical downscaling uses years of observational data to establish statistical relationships between large-scale climate conditions and regional climate factors. These relationships are then validated using independent observational data. These relationships are then applied to large-scale climate information output by general circulation models (GCMs) to predict future climate change scenarios for the region.

[0056] The steps of building a downscaling model are as follows: Figure 2 As shown:

[0057] S111: Obtain large-scale climate prediction factors and rainfall in the Xiaaoqiao area observed over many years, and establish a downscaling model using dynamic or statistical methods;

[0058] Establish a statistical function relationship between large-scale climate prediction factors and regional climate prediction variables:

[0059] Y=F(X)

[0060] In the formula, X is a large-scale climate forecast factor, such as air humidity, air temperature, atmospheric pressure and other climate conditions that can be measured by the weather monitoring station installed in the concave bridge; Y is a regional climate forecast variable, which in the present invention is the rainfall in the concave bridge area during a certain period of time (such as 5-minute rainfall, 30-minute rainfall, 6-hour rainfall, daily rainfall, etc.); F is a statistical relationship between the established large-scale climate forecast factor and the rainfall in the concave bridge area, which is generally obtained through dynamic methods (regional climate model simulation) or statistical methods (determination by observational data). This relationship is to show that one or more factors can be found to be related to rainfall through mathematical methods. In this formula, the large-scale forecast factor is a parameter related to the rainfall in the concave bridge area. The large-scale forecast factor can be used to determine the possibility of rainfall and the amount of rainfall.

[0061] S112: Validate the downscaling model.

[0062] From all large-scale climate predictors, select predictor variables suitable for analyzing precipitation in Xiaaoqiao. It should be noted that predictor variables can be obtained from meteorological monitoring stations, and the selected predictor variables must be recognizable and simulated by the global atmospheric circulation model. Furthermore, there must be no mutual influence between the selected predictor variables. This means that the correlation between the selected predictor variables is low. If the correlation between the predictor variables is high, one set of predictor variable data can be used to calculate the other set of predictor variables, effectively equivalent to inputting only one set of predictor variables.

[0063] Based on the actual data of the predictor variables observed during the calibration period and the downscaling model, a multivariate linear regression equation system is established.

[0064] Based on the precision of the warning time, an appropriate calibration period is selected, which is generally 3 to 5 years. The actual observation data of the forecast factor variables within the calibration period are used. The forecast factor variables are processed by stepwise linear regression analysis to establish a multivariate linear regression equation system.

[0065] Among them, multiple linear regression refers to the process of substituting one or more predictor variables into the above-mentioned downscaling model in the regression analysis of statistical downscaling to form a set of multiple linear regression equations, so as to obtain the optimal combination of multiple predictor variables to jointly predict or estimate the predicted variable.

[0066]

[0067] Where w is the probability of precipitation occurring in a certain period; a is the regression coefficient obtained by the linear least squares method through the multidimensional equation system consisting of the forecast factor variables and rainfall; u is the selected forecast factor, and t-1 represents the previous period.

[0068] By comparing with the uniformly distributed random number table r (0≤r≤1), it is judged whether rainfall occurs. t ≤r t It is time for rain.

[0069] The amount of rainfall is reflected by Z:

[0070]

[0071] Where Z is the Z score for a certain day; b is the regression coefficient obtained by the linear least squares method through the multidimensional equation system consisting of the forecast factor variables and rainfall.

[0072] The rainfall on day t is:

[0073] Y t =F -1 [(fZ t )]

[0074] Where f is the cumulative normal distribution function; F is the empirical distribution function of daily precipitation Yt.

[0075] The rainfall simulation is obtained based on the actual data of the predictor variables observed during the validation period and the multivariate linear regression equations.

[0076] Based on the calibration period, select an appropriate validation period, typically 10 to 20 months. Actual data from three or more weather monitoring stations at Xiaaoqiao during the validation period are collected and substituted into the established multivariate linear regression system to invert the precipitation time series for the validation period. Statistically analyze the various precipitation indicators and calculate the rainfall amount, which serves as the simulated rainfall. Compare this with the observed values ​​from the validation period to determine the accuracy of the multivariate linear regression system. If the regression equation differs significantly from the observed values, select new appropriate predictors and repeat the above steps.

[0077] The downscaling model was tested using statistical correlation coefficients based on the observed rainfall and simulated rainfall during the validation period.

[0078] Among them, the statistical correlation coefficient includes the correlation coefficient and the Nash coefficient. The correlation coefficient R 2 The downscaling model is tested simultaneously with the two coefficients R and Nash coefficient. If both meet the requirements, it can be considered that the downscaling model can well predict the rainfall in the Xiaaoqiao area. 2 It can be obtained by linear regression method to evaluate the degree of data consistency between measured values ​​and simulated values:

[0079] (1)R 2 =1 is very consistent, R 2 When <1, the smaller the value, the lower the degree of data fit.

[0080] (2) Nash coefficient, Nash-Suttcliffe coefficient.

[0081]

[0082] Where Q oi is the measured value at time i; Q pi is the simulated value obtained using the statistical downscaling method at time i; Q avg is the measured average value; n is the number of measurements. oi =Q pi When E ns =1, if E ns A negative value indicates that the model simulation average value is less reliable than the measured average value. ns A value above 0.6 is considered reliable.

[0083] After obtaining the downscaling model, the data of large-scale climate factors obtained by the weather monitoring station near the Xiasao Bridge were substituted into the downscaling model to obtain the rainfall in the Xiasao Bridge area.

[0084] S20: Divide the sunken bridge area into a main catchment area and multiple secondary catchment areas, and obtain elevation data of the main catchment area and each secondary catchment area.

[0085] The demarcation criteria are primarily based on the actual topography of the sunken bridge area. The primary catchment area refers to the area within the sunken bridge area where the low-lying water system and approach road pavement can collect water or receive rainfall. The secondary catchment area refers to areas that do not directly cover the primary study area but where surface runoff from precipitation may flow into the primary catchment area. After demarcation, elevation data for the primary catchment area and each secondary catchment area is obtained. This elevation data is primarily obtained using urban DEM data or field measurements. DEM stands for Digital Elevation Model, which is used to obtain this elevation data.

[0086] S30: Determine the risk rating of each sub-catchment based on the rainfall in the concave bridge area, the elevation of the main catchment area, and the elevation data of each sub-catchment area.

[0087] When the elevation data of the main catchment area is greater than that of the secondary catchment area, the formula is:

[0088]

[0089] Determine the risk factor of the catchment area, where K is the risk factor, R is the rainfall in the concave bridge area, and H is the 主 The elevation data of the main catchment area, H 次 It is the elevation data of the secondary catchment area.

[0090] Determine the risk rating based on the risk factor and the preset correspondence.

[0091] The correspondence between the risk coefficient and the risk rating is preset and is specifically reflected by the preset correspondence. The preset correspondence is as follows:

[0092] When the risk coefficient is greater than the first preset value, the risk rating is level one, indicating that there is a great risk of cross-catchment water collection. The first preset value ranges from 0.9 to 1.1, preferably 1, that is, when K>1, the risk rating of the secondary catchment is level one.

[0093] When the risk coefficient is less than or equal to the first preset value and greater than the second preset value, the risk rating is the second level, where the first preset value ranges from 0.7 to 0.9, preferably 0.8, that is, when 1≥K>0.8, the risk rating of the secondary catchment area is the second level.

[0094] When the risk coefficient is less than or equal to the second preset value and greater than the third preset value, the risk rating is the third level, wherein the third preset value ranges from 0.4 to 0.7, preferably 0.5, that is, when 0.8≥K>0.5, the risk rating of the secondary catchment area is the third level.

[0095] When the risk coefficient is less than or equal to the third preset value, the risk rating is level four, that is, when K≤0.5, the risk rating of the secondary catchment area is level four.

[0096] It should be noted that when the elevation data of the main catchment area is lower than that of the secondary catchment area, the risk rating of the secondary catchment area is level one.

[0097] S40: According to the risk rating of each sub-catchment area, the amount of water flowing from each sub-catchment area into the main catchment area is obtained.

[0098] When the risk rating is level 1, the amount of water flowing from the secondary catchment into the main catchment is equal to the runoff generated by the underlying surface of the secondary catchment;

[0099] When the risk rating is level 2, the water volume of the secondary catchment area flowing into the primary catchment area is equal to the runoff volume generated by the underlying surface of the secondary catchment area multiplied by a first preset multiple, wherein the first preset multiple ranges from 0.2 to 0.4, and is preferably 0.3;

[0100] When the risk rating is level three, the water volume of the secondary catchment area flowing into the main catchment area is equal to the runoff volume generated by the underlying surface of the secondary catchment area multiplied by a second preset multiple, wherein the second preset multiple ranges from 0.05 to 0.2, and is preferably 0.1;

[0101] When the risk rating is level 4, the secondary catchment area does not feed water into the main catchment area.

[0102] S50: Obtain the total accumulated water volume of the main catchment area according to the amount of water collected by each secondary catchment area into the main catchment area and the accumulated water volume of the main catchment area itself.

[0103] It should be noted that the runoff generated by the underlying surface of the secondary catchment area and the water accumulation in the primary catchment area itself are both obtained through the urban sunken bridge waterlogging model.

[0104] Urban sunken bridge waterlogging models are typically based on software such as ICMinfoworks, DHIMIKE, SWMM, and SimuWater. These models incorporate various monitoring data, including the sunken bridge catchment elevation data (DEM), rainfall, rainfall intensity, rainfall duration, underlying surface runoff, infiltration, and drainage. Using SimuWater as an example, a forecasting and early warning model for urban sunken bridge waterlogging was developed. This model primarily predicts future changes in waterlogging depth in the main catchment area of ​​urban sunken bridges, and also predicts the depth and rate of waterlogging change before rainfall occurs.

[0105] Finally, it should be noted that the above specific implementation methods are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to examples, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for predicting waterlogging at sunken bridges based on downscaling and dynamic division of catchment areas, characterized by: include: S10: Predict the rainfall in the area of ​​the sunken bridge; S20: Divide the sunken bridge area into a main catchment area and multiple secondary catchments, and obtain elevation data of the main catchment area and each secondary catchment area; S30: Determine the risk rating of each sub-catchment based on the rainfall in the Sag Bridge area, the elevation of the main catchment area, and the elevation data of each sub-catchment area; S40: according to the risk rating of each sub-catchment area, obtain the water volume of each sub-catchment area into the main catchment area; The preset multiple is determined based on the risk rating, and the runoff volume generated by the underlying surface of the secondary catchment area is obtained based on the preset multiple and the runoff volume generated by the underlying surface of the secondary catchment area. S50: Obtain the total accumulated water volume of the main catchment area according to the amount of water collected by each secondary catchment area into the main catchment area and the accumulated water volume of the main catchment area itself.

2. The method for predicting waterlogging at sunken bridges based on downscaling and dynamic watershed division according to claim 1, characterized in that: The step S10 includes: S110: Establishing a downscaling model by using a statistical downscaling method, wherein the downscaling model is a statistical relationship between large-scale climate prediction factors and rainfall in the Xiasao Bridge area; S120: Predict rainfall in the Xiaaoqiao area based on the downscaling model and large-scale climate prediction factors.

3. The method for predicting waterlogging at sunken bridges based on downscaling and dynamic watershed division according to claim 2, characterized in that: The step S110 includes: S111: Obtain large-scale climate prediction factors and rainfall in the Xiaaoqiao area observed over many years, and establish a downscaling model using dynamic or statistical methods; S112: Validate the downscaling model.

4. The method for predicting waterlogging at sunken bridges based on downscaling and dynamic watershed division according to claim 3, characterized in that: The step S112 includes: Select the predictor variables suitable for analyzing the precipitation in Xiasaoqiao from all the large-scale climate predictors; According to the actual data of the predictor variables observed during the calibration period and the downscaling model, a multivariate linear regression equation system is established; The rainfall simulation is obtained based on the actual data of the predictor variables observed during the validation period and the multivariate linear regression equation system; The downscaling model was tested using statistical correlation coefficients based on the observed rainfall and simulated rainfall during the validation period.

5. The method for predicting waterlogging at sunken bridges based on downscaling and dynamic watershed division according to claim 4, characterized in that: The statistical correlation coefficient includes a correlation coefficient and a Nash coefficient.

6. The method for predicting waterlogging at a sunken bridge based on downscaling and dynamic watershed division according to any one of claims 1 to 5, characterized in that: The large-scale climate prediction factors of the Xiaao Bridge area are obtained by the weather monitoring station built near the Xiaao Bridge.

7. The method for predicting waterlogging at sunken bridges based on downscaling and dynamic watershed division according to claim 1, characterized in that: The step S30 includes: When the elevation data of the main catchment area is greater than that of the secondary catchment area, the formula is: ， Determine the risk factor of the catchment area, where K is the risk factor, R is the rainfall in the concave bridge area, and H is the 主 The elevation data of the main catchment area, H 次 It is the elevation data of the secondary catchment area; Determine the risk rating based on the risk coefficient and the preset correspondence; When the elevation data of the main catchment area is lower than that of the secondary catchment area, the risk rating of the secondary catchment area is level one.

8. The method for predicting waterlogging at sunken bridges based on downscaling and dynamic watershed division according to claim 7, characterized in that: The preset corresponding relationship includes: When the risk factor is greater than the first preset value, the risk rating is level one; When the risk factor is less than or equal to the first preset value and greater than the second preset value, the risk rating is the second level; When the risk factor is less than or equal to the second preset value and greater than the third preset value, the risk rating is level three; When the risk coefficient is less than or equal to the third preset value, the risk rating is level four.

9. The method for predicting waterlogging at sunken bridges based on downscaling and dynamic watershed division according to claim 8, characterized in that: The step S40 includes: When the risk rating is level 1, the amount of water flowing into the main catchment from the sub-catchment is equal to the runoff generated by the underlying surface of the sub-catchment; When the risk rating is level 2, the amount of water flowing from the secondary catchment area into the main catchment area is equal to the runoff generated by the underlying surface of the secondary catchment area multiplied by the first preset multiple; When the risk rating is level three, the amount of water flowing from the secondary catchment area into the main catchment area is equal to the runoff generated by the underlying surface of the secondary catchment area multiplied by the second preset multiple; When the risk rating is level four, the secondary catchment area does not feed water into the main catchment area.

10. The method for predicting waterlogging at a sunken bridge based on downscaling and dynamic watershed division according to claim 9, characterized in that: The runoff generated by the underlying surface of the secondary catchment area and the water accumulation in the primary catchment area itself are both obtained through the urban sunken bridge waterlogging model.

Images