Flood risk prediction method and device based on artificial intelligence and emergence constraint

Abstract

The invention relates to the technical field of data prediction processing, in particular to a flood risk prediction method and device based on artificial intelligence and emergence constraint, and the method comprises the steps: segmenting a base flow through employing a preset digital filtering algorithm, and generating a surface runoff series of a research basin; obtaining a drainage basin average daily air temperature series, and dividing the same-day air temperature-surface runoff combination in the target research time period into a plurality of box elements with the same sample capacity by using a box element scaling function; separating a thermodynamic item and a dynamic item of the drainage basin extreme rainfall scale coefficient; predicting a preliminary prediction value of the extreme surface runoff driven by the thermal action; and constructing an emergence constraint model according to the data of the global climate mode and the predicted initial predicted value of the extreme surface runoff so as to predict the flood risk driven by the future thermal action of the target. Therefore, the problem that the flood risk under the driving of the future thermal action is predicted without comprehensively considering the atmospheric thermal power and power separation mechanism and without considering the emergence constraint in the related technology is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of data prediction and processing, and in particular to a flood risk prediction method and device based on artificial intelligence and emergence constraints. Background Art

[0002] In recent decades, the emission of greenhouse gases such as carbon dioxide in the atmosphere has increased dramatically, the energy balance and material circulation process of the earth's land and atmosphere have changed, the global temperature has continued to rise, coral reefs and tropical rainforests have died out on a large scale, and the imbalance and inequality of the social and economic systems and ecological environment have become increasingly serious. In order to meet the above challenges, it is urgent to study the formation mechanism and risk assessment technology of extreme hydrological events.

[0003] In the related technology, it is possible to collect meteorological and hydrological data sets and train multiple river basin hydrological models, establish a joint probability distribution function, and predict flood risks based on the joint probability distribution function and the weight parameters of each combined scenario.

[0004] However, the relevant technology does not comprehensively consider the atmospheric thermal and dynamic separation mechanism, and does not consider the emergence constraints to predict future flood risks driven by thermal effects, which urgently needs to be improved. Summary of the invention

[0005] The present invention provides a flood risk prediction method and device based on artificial intelligence and emergence constraints to solve the problem that the related technology does not comprehensively consider the atmospheric thermal and dynamic separation mechanism, and does not consider emergence constraints to predict future flood risks driven by thermal effects.

[0006] The first aspect of the present invention provides a flood risk prediction method based on artificial intelligence and emergence constraints, which is applied to the model construction stage and includes the following steps: obtaining daily runoff observation data at the outlet section of the basin, gridded atmospheric reanalysis data within the basin, and meteorological variables output by the global climate model; based on the daily runoff observation data, using a preset digital filtering algorithm to segment the base flow and generate a surface runoff series for the study basin; based on the surface runoff series for the study basin, inputting the temperature of the gridded atmospheric reanalysis data to obtain the average daily temperature series for the basin, and based on the average daily temperature series for the basin, using a box element scaling function to divide the temperature-surface runoff combination on the same day within the target study period into multiple samples with the same sample size. box elements to obtain the scaling coefficients of the extreme surface runoff and near-ground temperature in the study basin; according to the atmospheric column dry static energy balance and water balance equation and the scaling coefficients of the extreme surface runoff and the near-ground temperature, the thermal term and dynamic term of the extreme precipitation scaling coefficient of the basin are separated; based on the meteorological data output by the global climate model, the thermal term and dynamic term of the extreme precipitation scaling coefficient of the basin and the target temperature change value, the preliminary prediction value of the extreme surface runoff driven by thermal action is predicted to obtain the preliminary prediction value of the predicted extreme surface runoff; based on the data of the global climate model and the preliminary prediction value of the predicted extreme surface runoff, an emergence constraint model is constructed to predict the flood risk driven by the thermal action of the target in the future.

[0007] Optionally, in one embodiment of the present invention, the filtering equation in the preset digital filtering algorithm is: , in, for Surface runoff (surface runoff) after filtering during the period, for The original traffic volume during the period, For the The original traffic volume during the period, Characterize filter parameters; The calculation formula of the base flow is: , in, For the Surface runoff (surface runoff) after filtering during the period, For the The raw traffic volume during the time period.

[0008] Optionally, in one embodiment of the present invention, based on the average daily temperature series of the basin, a box scaling function is used to divide the temperature-surface runoff combination on the same day in the study period into multiple boxes with the same sample capacity to obtain the scaling coefficients of the extreme surface runoff and near-ground temperature of the study basin, including: using the average daily temperature of the basin as the sorting criterion, dividing the temperature-surface runoff combination on the same day in the study period into the multiple boxes with the same sample capacity to obtain the division result; based on the division result, using the median of the daily temperature to characterize the box temperature, and using the average intensity of the daily runoff that meets preset conditions to characterize the flood event; and obtaining the scaling coefficients of the extreme surface runoff and the near-ground temperature of the study basin according to the box temperature and the flood event.

[0009] Optionally, in one embodiment of the present invention, the calculation formula of the atmospheric column dry static energy balance is: , in, Characterizes the dry static energy, Characterize the time variable, Respectively characterize the mixing ratios of rain, cloud droplets, snow, graupel and cloud ice, represents the latent heat of sublimation, is the latent heat of vaporization, P To provide precipitation to the ground, Characterizes the mass-weighted integral in the vertical direction.

[0010] Optionally, in one embodiment of the present invention, separating the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the watershed includes: obtaining the air temperature and vertical wind speed series of different pressure layers averaged in the watershed to obtain the meteorological variables averaged in the watershed; and generating the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the watershed according to the meteorological variables averaged in the watershed.

[0011] Optionally, in one embodiment of the present invention, the calculation formula of the thermal term of the extreme precipitation scaling coefficient of the watershed is: , in, is the scaling factor of the extreme surface runoff, is the vertical wind speed, is the symbol of partial derivative, Indicates the amount of change, is the saturated water vapor pressure, p is the atmospheric pressure; The calculation formula of the thermal term of the predicted extreme surface runoff is: , in, is the extreme surface runoff in the historical period. is the thermal term of the scaling coefficient of the extreme surface runoff, is the average temperature change in the basin.

[0012] Optionally, in one embodiment of the present invention, the emergence constraint model is constructed based on the global climate model output and the preliminary predicted value of the predicted extreme surface runoff, including: calculating the changing trend of the average annual temperature in the historical period of each climate model based on the basin-average annual temperature of the global climate model; integrating the data of the global climate model to obtain a paired combination of the changing trend of the average annual temperature in the historical period and the target period; and constructing the emergence constraint model based on the paired combination and the preliminary predicted value of the predicted extreme surface runoff.

[0013] The second aspect of the present invention provides a flood risk prediction method based on artificial intelligence and emergence constraints, which is applied to the model application stage and includes the following steps: obtaining data from global climate models and preliminary prediction values ​​of predicted extreme surface runoff; inputting the data from the global climate model and the preliminary prediction values ​​of the predicted extreme surface runoff into a pre-constructed emergence constraint model to use the surface runoff in the emergence constraint model to characterize the flood risk; and constraining the prediction results of the flood risk, wherein the emergence constraint model is constructed by the data from the global climate model and the preliminary prediction values ​​of the predicted extreme surface runoff.

[0014] The third aspect of the present invention provides a flood risk prediction device based on artificial intelligence and emergence constraints, which is applied to the model construction stage, including: an acquisition module, which is used to obtain daily runoff observation data at the outlet section of the basin, gridded atmospheric reanalysis data within the basin, and meteorological variables output by the global climate model; a generation module, which is used to segment the base flow based on the daily runoff observation data using a preset digital filtering algorithm to generate a surface runoff series for the study basin; a division module, which is used to input the temperature of the gridded atmospheric reanalysis data based on the surface runoff series of the study basin to obtain the average daily temperature series of the basin, and based on the average daily temperature series of the basin, use a box element scaling function to divide the temperature-surface runoff combination on the same day in the target study period into multiple boxes with the same sample capacity. element, so as to obtain the scaling coefficients of the extreme surface runoff and near-ground temperature of the study basin; a separation module, used to separate the thermal term and dynamic term of the extreme precipitation scaling coefficient of the basin according to the atmospheric column dry static energy balance and water balance equation and the scaling coefficients of the extreme surface runoff and the near-ground temperature; a prediction module, used to predict the preliminary prediction value of the extreme surface runoff driven by thermal action based on the meteorological data output by the global climate model, the thermal term and dynamic term of the extreme precipitation scaling coefficient of the basin and the target temperature change value, and obtain the preliminary prediction value of the predicted extreme surface runoff; a construction module, used to construct an emergence constraint model based on the data of the global climate model and the preliminary prediction value of the predicted extreme surface runoff, so as to predict the flood risk driven by the thermal action of the target in the future.

[0015] Optionally, in one embodiment of the present invention, the filtering equation in the preset digital filtering algorithm is: , in, for Surface runoff (surface runoff) after filtering during the period, for The original traffic volume during the period, For the The original traffic volume during the period, Characterize filter parameters; The base flow is calculated as follows: , in, For the Surface runoff (surface runoff) after filtering during the period, For the The raw traffic volume during the time period.

[0016] Optionally, in one embodiment of the present invention, the division module includes: a division unit, used to divide the temperature-surface runoff combination on the same day in the study period into the multiple boxes with the same sample capacity based on the average daily temperature of the basin as the sorting criterion, to obtain a division result; a characterization unit, used to characterize the box temperature based on the division result using the median of the daily temperature, and to characterize the flood event using the average intensity of the daily runoff that meets preset conditions; a first generation unit, used to obtain the scaling coefficient of the extreme surface runoff and the near-ground temperature of the study basin based on the box temperature and the flood event.

[0017] Optionally, in one embodiment of the present invention, the calculation formula of the atmospheric column dry static energy balance is: , in, Characterizes the dry static energy, Characterize the time variable, Respectively characterize the mixing ratios of rain, cloud droplets, snow, graupel and cloud ice, represents the latent heat of sublimation, is the latent heat of vaporization, P To provide precipitation to the ground, Characterizes the mass-weighted integral in the vertical direction.

[0018] Optionally, in one embodiment of the present invention, the separation module includes: an acquisition unit, used to obtain the air temperature and vertical wind speed series of different pressure layers averaged in the watershed to obtain the meteorological variables averaged in the watershed; and a second generation unit, used to generate thermal terms and dynamic terms of the extreme precipitation scaling coefficient of the watershed based on the meteorological variables averaged in the watershed.

[0019] Optionally, in one embodiment of the present invention, the calculation formula of the thermal term of the extreme precipitation scaling coefficient of the watershed is: , in, is the scaling factor of the extreme surface runoff, is the vertical wind speed, is the symbol of partial derivative, Indicates the amount of change, is the saturated water vapor pressure, p is the atmospheric pressure; The calculation formula of the thermal term of the predicted extreme surface runoff is: , in, is the extreme surface runoff in the historical period. is the thermal term of the scaling coefficient of the extreme surface runoff, is the average temperature change in the basin.

[0020] Optionally, in one embodiment of the present invention, the construction module includes: a calculation unit, used to calculate the changing trend of the average annual temperature in the historical period of each climate model based on the average annual temperature of the watershed of the global climate model; a processing unit, used to integrate the data of the global climate model to obtain a paired combination of the changing trend of the average annual temperature in the historical period and the target period; and a construction unit, used to construct the emergence constraint model based on the paired combination and the preliminary predicted value of the predicted extreme surface runoff.

[0021] The fourth aspect of the present invention provides a flood risk prediction device based on artificial intelligence and emergence constraints, which is applied to the model application stage, including: a prediction value acquisition module, used to obtain data from the global climate model and preliminary prediction values ​​of the predicted extreme surface runoff; a characterization module, used to input the data from the global climate model and the preliminary prediction values ​​of the predicted extreme surface runoff into a pre-constructed emergence constraint model, so as to characterize the flood risk using the surface runoff in the emergence constraint model; a constraint module, used to constrain the prediction results of the flood risk, wherein the emergence constraint model is constructed by the data from the global climate model and the preliminary prediction values ​​of the predicted extreme surface runoff.

[0022] The embodiments of the present invention provide an important and highly operational reference basis for global and regional rainstorm disaster risk assessment and early warning under climate change scenarios by coupling the physical mechanism of flood risk changes and the statistical deviation correction method, and provide engineering reference value for responding to future climate disasters and scientifically formulating emission reduction strategies. This solves the problem that the relevant technologies do not comprehensively consider the atmospheric thermal and dynamic separation mechanism, and do not consider the emergence constraints to predict the flood risk driven by future thermal effects.

[0023] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which: Figure 1 A flow chart of a flood risk prediction method based on artificial intelligence and emergence constraints provided in an embodiment of the present invention applied to a model building stage; Figure 2 A flow chart of a flood risk prediction method based on artificial intelligence and emergence constraints provided in an embodiment of the present invention applied to a model application stage; Figure 3A schematic structural diagram of a flood risk prediction device based on artificial intelligence and emergence constraints provided in an embodiment of the present invention applied to a model building stage; Figure 4 A structural schematic diagram of a flood risk prediction device based on artificial intelligence and emergence constraints provided according to an embodiment of the present invention applied to a model application stage. DETAILED DESCRIPTION

[0025] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0026] The following describes a flood risk prediction method and device based on artificial intelligence and emergence constraints according to an embodiment of the present invention with reference to the accompanying drawings. In view of the problem that the related technologies mentioned in the above background technology do not comprehensively consider the atmospheric thermal and power separation mechanism, and do not consider the emergence constraints to predict the flood risk driven by future thermal effects, the present invention provides a flood risk prediction method based on artificial intelligence and emergence constraints. In this method, by coupling the physical mechanism of flood risk changes and the statistical deviation correction method, an important and highly operational reference basis is provided for global and regional rainstorm disaster risk assessment and early warning under climate change scenarios, and an engineering reference value is provided for responding to future climate disasters and scientifically formulating emission reduction strategies. As a result, the problem that the related technologies do not comprehensively consider the atmospheric thermal and power separation mechanism, and do not consider the emergence constraints to predict the flood risk driven by future thermal effects is solved.

[0027] Specifically, Figure 1 A schematic flow chart of a flood risk prediction method based on artificial intelligence and emergence constraints provided in an embodiment of the present invention.

[0028] like Figure 1 As shown in the figure, the flood risk prediction method based on artificial intelligence and emergence constraints is applied in the model building stage and includes the following steps: In step S101, daily runoff observation data of the outlet section of the basin, gridded atmospheric reanalysis data within the basin, and meteorological variables output by the global climate model are obtained.

[0029] In the actual implementation process, the embodiment of the present invention can obtain the daily runoff observation data of the outlet section of the study basin, and then collect the daily 2m (meter) air temperature of each grid point within the study basin from 1985 to 2014 from the fifth generation atmospheric reanalysis data set (ERA5) of the European Center for Medium-Range Weather Forecasts. In order to quantitatively separate the atmospheric thermal and dynamic effects, the hourly air temperature and vertical wind speed data of 20 layers (50hPa interval) from 50hPa to 1000hPa of the ERA5 data set are used and converted into daily scale data with a spatial resolution of 0.5°×0.5°.

[0030] In order to predict future climate scenarios, 30 global climate models newly released by the Coupled Model Inter-comparison Project Phase 6 (CMIP6) are used. Compared with the earlier version (CMIP5), CMIP6 uses the matrix framework of Shared Socioeconomic Pathway (SSP) and Representative Concentration Pathway (RCP). The embodiment of the present invention uses the model output data of three shared socioeconomic pathways, namely SSP126, SSP370 and SSP585. The output variables of the 30 global climate models include daily precipitation and temperature data. 1985-2014 (30 years in total) is selected as the historical period. In order to predict the meteorological and hydrological characteristics under the future climate change scenario, the last 30-year window of this century (2071-2100) is set as the future period.

[0031] In step S102, based on the daily runoff observation data, a preset digital filtering algorithm is used to segment the base flow and generate a surface runoff series for the study basin.

[0032] In the actual implementation process, the embodiment of the present invention can use a digital filtering algorithm to segment the base flow based on the daily runoff observation data of the study basin to obtain the surface runoff series of the study basin.

[0033] Among them, in one embodiment of the present invention, the digital filtering method is a method for segmenting the base stream that has been widely used in recent years. The embodiment of the present invention adopts the recursive digital filter proposed by Lyne and Hollick in 1979, and its filtering equation is: , (1) in, for Surface runoff (surface runoff) after filtering during the period, for The original traffic volume during the period, for The original traffic volume during the period, Characterize the filter parameters.

[0034] After obtaining the total runoff and surface runoff, the base flow calculation formula is: , (2) in, for Surface runoff (surface runoff) after filtering during the period, for The raw traffic volume during the time period.

[0035] The outflow process of groundwater can be calculated by equation (1) and equation (2), so as to segment the flow process. The present invention sets the number of filtering times to 3, and obtains the surface runoff process by optimizing the solution through the EcoHydRology package based on the R program.

[0036] In step S103, based on the surface runoff series of the study basin, the temperature of the gridded atmospheric reanalysis data is input to obtain the basin average daily temperature series, and based on the basin average daily temperature series, the temperature-surface runoff combination on the same day in the target study period is divided into multiple bins with the same sample capacity using a bin scaling function to obtain the scaling coefficients of the extreme surface runoff and near-ground temperature in the study basin.

[0037] It is understandable that the Binning scaling function can better characterize the nonlinear relationship between variables and has been widely used in recent years to analyze the response characteristics of extreme precipitation and runoff to global warming.

[0038] In the actual implementation process, the embodiment of the present invention can adopt the Thiessen polygon method, take the 2m temperature of the ERA5 reanalysis data set as input, obtain the basin-average daily temperature series, and based on the basin-average daily temperature series, use the box element scaling function to divide the temperature-surface runoff combination on the same day in the target research period into 12 boxes with the same sample capacity, so as to obtain the scaling coefficients of the extreme surface runoff and near-ground temperature in the study basin.

[0039] Optionally, in one embodiment of the present invention, based on the average daily temperature series of the basin, a box scaling function is used to divide the temperature-surface runoff combination on the same day in the study period into multiple boxes with the same sample capacity to obtain the scaling coefficients of the extreme surface runoff and near-ground temperature in the study basin, including: using the average daily temperature of the basin as the sorting criterion, dividing the temperature-surface runoff combination on the same day in the study period into multiple boxes with the same sample capacity to obtain the division results; based on the division results, using the median of the daily temperature to characterize the box temperature, and using the average intensity of the daily runoff that meets preset conditions to characterize the flood event; and obtaining the scaling coefficients of the extreme surface runoff and near-ground temperature in the study basin based on the box temperature and the flood event.

[0040] It can be understood that the average intensity of daily runoff that meets the preset conditions in the embodiment of the present invention can be the average intensity of three daily runoffs that are higher than the 99% quantile.

[0041] In the actual implementation process, the embodiment of the present invention can use the average daily temperature of the basin as the sorting criterion, divide the temperature-surface runoff combination series on the same day in the study period into 12 boxes with the same sample capacity, take the median of the daily temperature to represent the box temperature, and use the average intensity of three daily runoffs above the 99% quantile to represent the flood event, and finally obtain 12 groups of temperature-extreme surface runoff series.

[0042] The Clausius-Clapeyron (CC) equation can quantitatively describe the saturated water vapor pressure With temperature T The nonlinear relationship: , (3) in, is the latent heat of vaporization, which is 2.5×10 6 J kg -1 ; R v is the water vapor gas constant, which is 461 J kg -1 K -1 .

[0043] Atmospheric water vapor pressure is the material basis for the formation of precipitation and changes with temperature. T The increase shows a nonlinear growth; extreme surface runoff is mainly affected by extreme precipitation, so the following formula is used to measure the effect of extreme surface runoff on temperature variation: The response function is: , (4) in, P 1 and P 2 are the extreme surface runoff of adjacent boxes, Scaling rate, also known as response intensity.

[0044] The local weighted sliding regression smoothing method was used to fit equation (4). Previous studies have found that the climate response types of extreme surface runoff can be divided into monotonic rising type, monotonic falling type and Hook structure (the inflection point is defined as the peak temperature, T pp ). The present invention first identifies the climate response type and T pp , and then the scaling coefficient is derived by the least squares method.

[0045] It should be noted that in the process of deriving the scaling coefficient, only the temperature below T pp Air temperature-surface runoff data in the box elements.

[0046] In step S104, the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the watershed are separated according to the atmospheric column dry static energy budget and water balance equation and the scaling coefficients of extreme surface runoff and near-surface temperature.

[0047] In actual implementation, the embodiment of the present invention can separate the thermal term and dynamic term of the extreme precipitation scaling coefficient of the basin based on the atmospheric column dry static energy budget and water balance equation and the scaling coefficients of extreme surface runoff and near-ground temperature.

[0048] Among them, in one embodiment of the present invention, the calculation formula of the atmospheric column dry static energy balance is: , (5) in, Characterizes the dry static energy, Characterize the time variable, Respectively characterize the mixing ratios of rain, cloud droplets, snow, graupel and cloud ice, represents the latent heat of sublimation, is the latent heat of vaporization, P To provide precipitation to the ground, Characterizes the mass-weighted integral in the vertical direction.

[0049] Extreme precipitation events are usually accompanied by strong atmospheric upward motion, so the dry static energy can be approximately represented by the vertical advection term, that is, ,in is the vertical wind speed; assuming that the average atmospheric lapse rate is close to the wet adiabatic process, the energy conservation law gives: , (6) in, is the latent heat of vaporization, represents saturated specific humidity, which is determined by air temperature and pressure to obtain; , is the ratio of the gas constants of dry air and water vapor, taken as 0.622.

[0050] neglect and The difference between is the extreme precipitation intensity, and the combined equations (5) and (6) yield: , (7) Defining precipitation efficiency considering water balance , formula (5) can be simplified as: , (8) in, , is the vertical wind speed, is the symbol of partial derivative, is the saturated water vapor pressure.

[0051] In order to study the different impact mechanisms of thermal and dynamic effects of the climate system on extreme precipitation, the change of precipitation efficiency is often ignored, and the following relationship is obtained using atmospheric static balance: , (9) in, Indicates the amount of change, For extreme precipitation intensity, is the vertical wind speed, is the symbol of partial derivative, p is the atmospheric pressure, is the saturated water vapor pressure.

[0052] Formula (9) can be further decomposed into: , (10) in, Indicates the amount of change, For extreme precipitation intensity, is the vertical wind speed, is the symbol of partial derivative, p is the atmospheric pressure, is the saturated water vapor pressure.

[0053] The atmospheric thermal effect affects the saturated water vapor pressure, which is only controlled by the temperature and pressure; the atmospheric dynamic effect affects the circulation intensity by changing the vertical wind speed, so the formula (10) can be decomposed into the thermal term and the dynamic term of extreme precipitation. The present invention uses the ERA5 data set 50hPa-1000hPa, a total of 20 layers of data, to calculate the vertical integral of formulas (7)-(10).

[0054] The Thiessen multi-deformation method is used to obtain the basin-averaged air temperature and vertical wind speed series of different pressure layers. The basin-averaged meteorological variables are then substituted into the above formula to obtain the thermal and dynamic terms of the scaling coefficient of extreme precipitation in the study basin.

[0055] Optionally, in one embodiment of the present invention, the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the watershed are separated, including: obtaining the air temperature and vertical wind speed series of different pressure layers averaged in the watershed to obtain the meteorological variables averaged in the watershed; and generating the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the watershed according to the meteorological variables averaged in the watershed.

[0056] In the actual implementation process, the embodiment of the present invention can adopt the Thiessen multi-deformation method to obtain the air temperature and vertical wind speed series of different pressure layers averaged in the basin, obtain the meteorological variables averaged in the basin, and generate the thermal term and dynamic term of the extreme precipitation scaling coefficient of the basin according to the meteorological variables averaged in the basin, thereby providing support for subsequent research on the thermal term of extreme surface runoff in the basin in the future.

[0057] In step S105, based on the meteorological data output by the global climate model, the thermal term and dynamic term of the extreme precipitation scaling coefficient of the basin and the target temperature change value, the preliminary prediction value of the extreme surface runoff driven by thermal action is predicted to obtain the preliminary prediction value of the predicted extreme surface runoff.

[0058] It can be understood that the scaling coefficient of the above-mentioned extreme surface runoff in the embodiment of the present invention is Rate. Since extreme surface runoff is mainly affected by extreme precipitation, it is assumed that the ratio of the thermal term and the dynamic term of the extreme precipitation scaling coefficient is still suitable for measuring the response characteristics of extreme surface runoff.

[0059] In actual implementation, the embodiment of the present invention can predict the preliminary predicted value of extreme surface runoff driven by thermal effects based on the meteorological data output by the global climate model, the thermal term of the scaling coefficient and the future temperature change value.

[0060] In one embodiment of the present invention, the calculation formula of the thermal term (Rate1) of the extreme precipitation scaling coefficient of the watershed is: , (11) in, is the scaling factor of extreme surface runoff, is the vertical wind speed, is the symbol of partial derivative, Indicates the amount of change, is the saturated water vapor pressure, p is the atmospheric pressure.

[0061] At each grid point in the study basin, for each global climate model under each shared socioeconomic path, the temperature change value in the future period relative to the historical period is calculated; further, the Thiessen polygon method is used to derive the average temperature change value of the basin, which is recorded as Δ T .

[0062] The 99% quantile of all daily surface runoff series from 1985 to 2014 is used to represent the flood magnitude in the historical period, denoted as ). Since the atmospheric thermal effect caused by rising temperature is an important mechanism driving the growth of extreme surface runoff, Δ T The thermal terms for predicting future extreme surface runoff are as follows: , (12) in, is the extreme surface runoff in the historical period. is the thermal term of the scaling coefficient of extreme surface runoff, is the average temperature change in the basin.

[0063] Using the above method, the thermal term of extreme surface runoff in the future period can be obtained for each global climate model under each shared socioeconomic pathway.

[0064] In step S106, an emergence constraint model is constructed based on the data of the global climate model and the preliminary predicted value of the predicted extreme surface runoff to predict the flood risk driven by the target future thermal effect.

[0065] As a possible implementation method, the embodiment of the present invention can construct an emergence constraint model based on the data of the global climate model and the preliminary predicted value of the predicted extreme surface runoff, use surface runoff to characterize flood risk, constrain the uncertainty of future flood risk prediction, and provide an important and highly operational reference basis for global and regional rainstorm disaster risk assessment and early warning under the climate change scenario, and provide engineering reference value for responding to future climate disasters and scientifically formulating emission reduction strategies.

[0066] Optionally, in one embodiment of the present invention, an emergence constraint model is constructed based on the global climate model output and the preliminary predicted values ​​of extreme surface runoff after prediction, including: calculating the changing trend of the average annual temperature in the historical period of each climate model based on the basin-average annual temperature of the global climate model; integrating the data of the global climate model to obtain a paired combination of the changing trend of the average annual temperature in the historical period and the target period; and constructing the emergence constraint model based on the paired combination and the preliminary predicted values ​​of extreme surface runoff after prediction.

[0067] In the actual implementation process, the embodiment of the present invention can use the Thiessen polygon method to deduce the average annual temperature of the watershed of each global climate model under each shared socio-economic path, and further calculate the change trend of the historical average annual temperature of each climate model (denoted as Furthermore, under each shared socioeconomic path, the data of 30 global climate models were integrated to obtain 30 sets of historical average annual temperature change trends and future periods. The emergence constraint model constructed by the pairing combination is as follows: , (13) in, a and b Represents the parameters of the emergence constraint model.

[0068] The least squares method is used to solve the parameters of the emergence constraint model. a and b Based on the reanalysis data of the ERA5 dataset, the trend of the annual average 2m temperature during 1985-2014 was deduced, and the trend term was substituted into the constructed emergence constraint model, as follows: , (14) in, represents the forecast results of the change of flood risk driven by thermal effects in the future period after correction, Represents the average annual temperature trend obtained from the ERA5 dataset at this grid point.

[0069] The emergent constraint models for each shared socioeconomic pathway were constructed separately, and finally the prediction results of three future flood risk changes were obtained, which can further evaluate the uncertainty range of future flood risk changes.

[0070] Further, Figure 2 A schematic flow chart of a flood risk prediction method based on artificial intelligence and emergence constraints provided in an embodiment of the present invention applied to a model application stage.

[0071] like Figure 2 As shown in the figure, the flood risk prediction method based on artificial intelligence and emergence constraints is applied in the model building stage and includes the following steps: In step S201, data of the global climate model and a preliminary predicted value of the predicted extreme surface runoff are obtained.

[0072] Among them, the embodiment of the present invention can predict the preliminary predicted value of extreme surface runoff driven by thermal effects based on the meteorological data output by the global climate model, the thermal term of the scaling coefficient and the future temperature change value.

[0073] In step S202, the data of the global climate model and the preliminary predicted value of the predicted extreme surface runoff are input into the pre-built emergence constraint model to characterize the flood risk using the surface runoff in the emergence constraint model.

[0074] Among them, the embodiment of the present invention can construct an emergence constraint model based on the data of the global climate model and the preliminary predicted value of the predicted extreme surface runoff, use surface runoff to characterize flood risk, constrain the uncertainty of future flood risk prediction, and provide an important and highly operational reference basis for global and regional rainstorm disaster risk assessment and early warning under the climate change scenario, and provide engineering reference value for responding to future climate disasters and scientifically formulating emission reduction strategies.

[0075] In step S203, the flood risk prediction results are constrained, wherein an emergence constraint model is constructed by using data from global climate models and preliminary prediction values ​​of predicted extreme surface runoff.

[0076] Among them, the embodiments of the present invention can constrain the prediction results of flood risks, and further constrain the uncertainty of future flood risk predictions.

[0077] According to the flood risk prediction method based on artificial intelligence and emergence constraints proposed in the embodiment of the present invention, by coupling the physical mechanism of flood risk changes and the statistical deviation correction method, it provides an important and highly operational reference basis for global and regional rainstorm disaster risk assessment and early warning under the climate change scenario, and provides engineering reference value for responding to future climate disasters and scientifically formulating emission reduction strategies. Thus, the problem that the relevant technology does not comprehensively consider the atmospheric thermal and dynamic separation mechanism, and does not consider the emergence constraints to predict the flood risk driven by future thermal effects is solved.

[0078] Next, a flood risk prediction device based on artificial intelligence and emergence constraints proposed according to an embodiment of the present invention is described with reference to the accompanying drawings.

[0079] Figure 3 It is a structural schematic diagram of a flood risk prediction device based on artificial intelligence and emergence constraints according to an embodiment of the present invention.

[0080] like Figure 3 As shown, the flood risk prediction device 10 based on artificial intelligence and emergence constraints is applied to the model construction stage and includes: an acquisition module 100, a generation module 200, a division module 300, a separation module 400, a prediction module 500 and a construction module 600.

[0081] Specifically, the acquisition module 100 is used to obtain daily runoff observation data at the outlet section of the basin, gridded atmospheric reanalysis data within the basin, and meteorological variables output by the global climate model.

[0082] The generation module 200 is used to segment the base flow based on the daily runoff observation data and generate the surface runoff series of the study basin by using a preset digital filtering algorithm.

[0083] The partitioning module 300 is used to input the temperature of the gridded atmospheric reanalysis data based on the surface runoff series of the study basin to obtain the basin average daily temperature series, and based on the basin average daily temperature series, use the box element scaling function to divide the temperature-surface runoff combination on the same day in the target study period into multiple boxes with the same sample capacity, so as to obtain the scaling coefficients of the extreme surface runoff and near-ground temperature of the study basin.

[0084] The separation module 400 is used to separate the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the basin according to the atmospheric column dry static energy budget and the moisture balance equation and the scaling coefficients of the extreme surface runoff and the near-surface temperature.

[0085] The prediction module 500 is used to predict the preliminary prediction value of extreme surface runoff driven by thermal effects based on the meteorological data output by the global climate model, the thermal term and dynamic term of the extreme precipitation scaling coefficient of the basin, and the target temperature change value, so as to obtain the preliminary prediction value of the predicted extreme surface runoff.

[0086] The construction module 600 is used to construct an emergence constraint model based on the data of the global climate model and the preliminary predicted value of the predicted extreme surface runoff to predict the flood risk driven by the target future thermal effect.

[0087] Optionally, in one embodiment of the present invention, the filtering equation in the preset digital filtering algorithm is: , in, for Surface runoff (surface runoff) after filtering during the period, for The original traffic volume during the period, for The original traffic volume during the period, Characterize filter parameters; The base flow is calculated as: , in, for Surface runoff (surface runoff) after filtering during the period, for The raw traffic volume during the time period.

[0088] Optionally, in one embodiment of the present invention, the division module 300 includes: a division unit, a characterization unit and a first generation unit.

[0089] Among them, the division unit is used to divide the temperature-surface runoff combination of the same day in the study period into multiple boxes with the same sample capacity based on the average daily temperature of the basin as the sorting criterion to obtain the division result.

[0090] The characterization unit is used to characterize the box temperature based on the division results, using the median of the daily temperature, and to characterize the flood event using the average intensity of the daily runoff that meets the preset conditions.

[0091] The first generation unit is used to obtain the scaling coefficients of extreme surface runoff and near-surface air temperature in the study basin based on the box temperature and flood events.

[0092] Optionally, in one embodiment of the present invention, the calculation formula of the atmospheric column dry static energy balance is: , in, Characterizes the dry static energy, t Characterize the time variable, Characterize the mixing ratios of rain, cloud droplets, snow, graupel and cloud ice respectively; represents the latent heat of sublimation, is the latent heat of vaporization, P To provide precipitation to the ground, Characterizes the mass-weighted integral in the vertical direction.

[0093] Optionally, in one embodiment of the present invention, the separation module 400 includes: an acquisition unit and a second generation unit.

[0094] The acquisition unit is used to obtain the average air temperature and vertical wind speed series of different pressure layers in the basin, and obtain the average meteorological variables in the basin.

[0095] The second generating unit is used to generate the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the basin according to the averaged meteorological variables of the basin.

[0096] Optionally, in one embodiment of the present invention, the calculation formula of the thermal term of the extreme precipitation scaling coefficient of the watershed is: , in, is the scaling factor of extreme surface runoff, is the vertical wind speed, is the symbol of partial derivative, Indicates the amount of change, is the saturated water vapor pressure, p is the atmospheric pressure; The calculation formula of the thermal term of the predicted extreme surface runoff is: , in, is the extreme surface runoff in the historical period. is the thermal term of the scaling coefficient of extreme surface runoff, is the average temperature change in the basin.

[0097] Optionally, in one embodiment of the present invention, the construction module 600 includes: a computing unit, a processing unit and a construction unit.

[0098] Among them, the calculation unit is used to calculate the changing trend of the historical average annual temperature of each climate model based on the average annual temperature of the river basin in the global climate model.

[0099] The processing unit is used to integrate the data of the global climate model to obtain the matching combination of the changing trend of the average annual temperature in the historical period and the target period.

[0100] A construction unit is used to construct an emergence constraint model based on the paired combination and the preliminary predicted values ​​of the predicted extreme surface runoff.

[0101] Figure 4 It is a structural schematic diagram of a flood risk prediction device based on artificial intelligence and emergence constraints according to an embodiment of the present invention.

[0102] like Figure 4 As shown, the flood risk prediction device 20 based on artificial intelligence and emergence constraints is applied in the model application stage and includes: a prediction value acquisition module 700, a characterization module 800 and a constraint module 900.

[0103] Specifically, the prediction value acquisition module 700 is used to obtain data of global climate models and preliminary prediction values ​​of predicted extreme surface runoff.

[0104] The characterization module 800 is used to input the data of the global climate model and the preliminary predicted value of the predicted extreme surface runoff into the pre-built emergence constraint model, so as to characterize the flood risk by using the surface runoff in the emergence constraint model.

[0105] The constraint module 900 is used to constrain the prediction results of flood risk, wherein the emergence constraint model is constructed by the data of the global climate model and the preliminary prediction value of the predicted extreme surface runoff.

[0106] It should be noted that the above explanation of an embodiment of a flood risk prediction method based on artificial intelligence and emergence constraints is also applicable to a flood risk prediction device based on artificial intelligence and emergence constraints in this embodiment, which will not be repeated here.

[0107] According to the embodiment of the present invention, a flood risk prediction device based on artificial intelligence and emergence constraints is proposed. By coupling the physical mechanism of flood risk changes and the statistical deviation correction method, it provides an important and highly operational reference basis for global and regional rainstorm disaster risk assessment and early warning under climate change scenarios, and provides engineering reference value for responding to future climate disasters and scientifically formulating emission reduction strategies. As a result, the problem that the relevant technology does not comprehensively consider the atmospheric thermal and dynamic separation mechanism, and does not consider the emergence constraints to predict the future flood risk driven by thermal effects is solved.

[0108] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0109] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "N" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0110] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present invention belong.

Claims

1. A flood risk prediction method based on artificial intelligence and emergence constraints, characterized in that: Applied to the model building phase, it includes the following steps: Obtain daily runoff observation data at the basin outlet section, gridded atmospheric reanalysis data within the basin, and meteorological variables output by the global climate model; Based on the daily runoff observation data, a preset digital filtering algorithm is used to segment the base flow and generate a surface runoff series for the study basin; Based on the surface runoff series of the study basin, the temperature of the gridded atmospheric reanalysis data is input to obtain the average daily temperature series of the basin, and based on the average daily temperature series of the basin, the temperature-surface runoff combination on the same day in the target study period is divided into multiple bins with the same sample capacity using a bin scaling function to obtain the scaling coefficients of the extreme surface runoff and near-ground temperature of the study basin; According to the atmospheric column dry static energy budget and water balance equation and the scaling coefficients of the extreme surface runoff and the near-surface temperature, the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the basin are separated; Based on the meteorological data output by the global climate model, the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the basin and the target temperature change value, a preliminary prediction value of extreme surface runoff driven by thermal action is predicted to obtain a preliminary prediction value of the predicted extreme surface runoff; An emergence constraint model is constructed based on the data of the global climate model and the preliminary predicted values ​​of the predicted extreme surface runoff to predict the flood risk driven by the thermal effects in the target future.

2. A flood risk prediction method based on artificial intelligence and emergence constraints according to claim 1, characterized in that: The filtering equation in the preset digital filtering algorithm is: ， in, for Surface runoff (surface runoff) after filtering during the period, for The original traffic volume during the period, For the The original traffic volume during the period, Characterize filter parameters; The calculation formula of the base flow is: ， in, For the Surface runoff (surface runoff) after filtering during the period, For the The raw traffic volume during the time period.

3. The flood risk prediction method based on artificial intelligence and emergence constraints according to claim 1 is characterized in that: Based on the average daily temperature series of the basin, the temperature-surface runoff combination on the same day in the study period is divided into multiple boxes with the same sample capacity using a box scaling function to obtain the scaling coefficients of the extreme surface runoff and near-surface temperature in the study basin, including: Taking the average daily temperature of the watershed as a sorting criterion, dividing the temperature-surface runoff combination on the same day within the study period into the multiple bins with the same sample capacity, and obtaining a division result; Based on the division results, the median of the daily air temperature is used to represent the box temperature, and the average intensity of the daily runoff that meets the preset conditions is used to represent the flood event; The scaling coefficients of the extreme surface runoff and the near-ground air temperature in the study basin are obtained according to the box temperature and the flood event.

4. A flood risk prediction method based on artificial intelligence and emergence constraints according to claim 1, characterized in that: The calculation formula of the atmospheric column dry static energy balance is: ， in, Characterizes the dry static energy, Characterize the time variable, Respectively characterize the mixing ratios of rain, cloud droplets, snow, graupel and cloud ice, represents the latent heat of sublimation, is the latent heat of vaporization, P To provide precipitation to the ground, Characterizes the mass-weighted integral in the vertical direction.

5. The flood risk prediction method based on artificial intelligence and emergence constraints according to claim 1 is characterized in that: The thermal and dynamic terms of the extreme precipitation scaling coefficient of the basin are separated, including: Obtain the average air temperature and vertical wind speed series of different pressure layers in the basin to obtain the basin-averaged meteorological variables; The thermal term and the dynamic term of the extreme precipitation scaling coefficient of the basin are generated according to the averaged meteorological variables of the basin.

6. A flood risk prediction method based on artificial intelligence and emergence constraints according to claim 1, characterized in that: The calculation formula of the thermal term of the extreme precipitation scaling coefficient of the basin is: ， in, is the scaling factor of the extreme surface runoff, is the vertical wind speed, is the symbol of partial derivative, Indicates the amount of change, is the saturated water vapor pressure, p is the atmospheric pressure; The calculation formula of the thermal term of the predicted extreme surface runoff is: ， in, is the extreme surface runoff in the historical period. is the thermal term of the scaling coefficient of the extreme surface runoff, is the average temperature change in the basin.

7. The flood risk prediction method based on artificial intelligence and emergence constraints according to claim 1 is characterized in that: The constructing of an emergence constraint model according to the global climate model output and the preliminary predicted value of the predicted extreme surface runoff comprises: Based on the average annual temperature of the watershed of the global climate model, calculate the change trend of the average annual temperature in the historical period of each climate model; Integrate and process the data of the global climate model to obtain a pairing combination of the change trend of the average annual temperature in the historical period and the target period; The emergence constraint model is constructed based on the paired combination and the preliminary predicted value of the predicted extreme surface runoff.

8. A flood risk prediction method based on artificial intelligence and emergence constraints, characterized in that: Applied to the model application phase, it includes the following steps: Obtain data from global climate models and preliminary estimates of projected extreme surface runoff; Inputting the data of the global climate model and the preliminary predicted value of the predicted extreme surface runoff into a pre-built emergence constraint model, so as to characterize the flood risk by using the surface runoff in the emergence constraint model; The flood risk prediction result is constrained, wherein the emergence constraint model is constructed by the data of the global climate model and the preliminary prediction value of the predicted extreme surface runoff.

9. A flood risk prediction device based on artificial intelligence and emergence constraints, characterized in that: Applied in the model building phase, including: The acquisition module is used to obtain the daily runoff observation data of the basin outlet section, the gridded atmospheric reanalysis data within the basin, and the meteorological variables output by the global climate model; A generation module, for segmenting the base flow based on the daily runoff observation data by using a preset digital filtering algorithm to generate a surface runoff series for the study basin; A partitioning module is used to input the temperature of the gridded atmospheric reanalysis data based on the surface runoff series of the study basin to obtain the average daily temperature series of the basin, and based on the average daily temperature series of the basin, use a bin scaling function to divide the temperature-surface runoff combination on the same day in the target study period into multiple bins with the same sample capacity, so as to obtain the scaling coefficients of the extreme surface runoff and near-ground temperature of the study basin; A separation module, for separating the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the watershed according to the atmospheric column dry static energy budget and the moisture balance equation and the scaling coefficients of the extreme surface runoff and the near-surface temperature; A prediction module, for predicting a preliminary prediction value of extreme surface runoff driven by thermal action based on the meteorological data output by the global climate model, the thermal term and the dynamic term of the extreme precipitation scaling coefficient of the basin, and the target temperature change value, to obtain a preliminary prediction value of the predicted extreme surface runoff; A construction module is used to construct an emergence constraint model based on the data of the global climate model and the preliminary predicted value of the predicted extreme surface runoff to predict the flood risk driven by the target future thermal effect.

10. A flood risk prediction device based on artificial intelligence and emergence constraints, characterized in that: Applied in the model application phase, including: A prediction value acquisition module is used to obtain data of global climate models and preliminary prediction values ​​of extreme surface runoff after prediction; A characterization module, for inputting the data of the global climate model and the preliminary predicted value of the predicted extreme surface runoff into a pre-built emergence constraint model, so as to characterize the flood risk by using the surface runoff in the emergence constraint model; A constraint module is used to constrain the prediction results of the flood risk, wherein the emergence constraint model is constructed by the data of the global climate model and the preliminary prediction value of the predicted extreme surface runoff.

Images