A physical mechanism-based method for transplanting spatial-temporal process of extreme rainstorm of a certain stadium

Abstract

The application provides a physical mechanism-based field extreme rainstorm spatiotemporal process transplantation method, which comprises the following steps: screening a WRF model suitable for the characteristics of a target region; constructing a similarity index system covering energy, water vapor and dynamic conditions, and screening historical rainstorm events with similar background fields as transplantation objects; decomposing meteorological data corresponding to the transplantation objects according to pressure layers, mapping the data to the target region through coordinate translation and rotation transformation, and replacing original boundary field parameters; combining the terrain characteristics of the target region, and performing physical constraint correction on the transformed meteorological parameters; and inputting the corrected parameters into the WRF model to simulate and generate target region extreme rainstorm spatiotemporal process data. By introducing background field similarity screening and multiple physical constraint correction, the application solves the problem that traditional rainstorm transplantation methods lack physical mechanism support, realizes scientific reconstruction of rainstorm processes with both historical intensity and local adaptability, and provides high-precision pre-visualization data support for flood control and disaster reduction.

Description

Technical Field

[0001] This invention relates to the fields of meteorology, hydrology, and flood control and disaster reduction, and in particular to a method for transplanting the spatiotemporal process of extreme rainstorms based on physical mechanisms. Background Technology

[0002] With the increasing frequency of extreme rainfall events under climate change conditions, it is crucial to conduct simulations of extreme rainfall events from other regions within the local area for disaster prevention and mitigation. To assess the potential impacts of extreme rainfall, it is often necessary to transplant historical extreme rainfall events from other locations into the local area for simulation analysis.

[0003] Existing methods for transplanting rainfall data mainly fall into two categories: direct translation and transposition. Direct translation involves directly moving historical rainfall data to the target area; transposition adjusts the spatial location of the rainfall process, taking into account changes in the rainfall center. However, the main drawback of these two methods is that they only perform a simple spatial displacement of rainfall data, neglecting the physical processes of rainfall initiation, formation, and development. This transplantation scenario ignores differences in physical conditions such as topography and circulation, resulting in a lack of physical mechanism support for the transplanted rainfall process in the target area. This makes it difficult to accurately reflect the actual probability of extreme rainfall occurring locally, effectively achieving only spatial displacement rather than a true transplantation of the physical processes. Summary of the Invention

[0004] This invention provides a method for transplanting the spatiotemporal process of extreme rainstorms based on physical mechanisms, in order to overcome the deficiencies in the existing technology.

[0005] This invention provides a method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms, comprising the following steps: Based on the underlying surface characteristics, climate background, and historical rainfall patterns of the target area, suitable WRF models are selected. Meteorological data of historical rainstorms are acquired, and a background field similarity index system including energy conditions, water vapor conditions and dynamic conditions is constructed. Based on the background field similarity index system, the comprehensive similarity index between the target area and each historical rainstorm area is calculated, and historical rainstorm areas with a comprehensive similarity index greater than the similarity judgment threshold are used as transplantation objects. Based on the preprocessed meteorological data of the transplanted object, meteorological parameters of each level are decomposed according to the key pressure layer. The meteorological parameters at each level of the transplanted object are transformed by coordinate transformation. The latitude and longitude coordinates of the historical rainstorm area where the transplanted object is located are translated to the center of the target area. The rotation is adjusted according to the topography and circulation characteristics of the target area to obtain the meteorological parameters at each level after coordinate transformation. The meteorological parameters at each level after coordinate transformation are used to replace the original atmospheric boundary field parameters at the corresponding level in the target area, and the meteorological parameters at each level after coordinate transformation are physically constrained and corrected according to the topographic features of the target area. The meteorological parameters at each level, after being corrected by physical constraints, are used as the initial and boundary field input parameters of the WRF model. The WRF model is then run to generate spatiotemporal process data of extreme rainstorms in the target area.

[0006] According to the present invention, a method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms is provided. The construction of a background field similarity index system comprising energy conditions, water vapor conditions, and dynamic conditions includes: The pressure layer temperature difference index was selected as the core index of energy conditions. Lower-level water vapor flux was selected as the core indicator of water vapor conditions. High-level geopotential height and low-level wind speed and direction are selected as the core indicators of dynamic conditions. The weighting coefficients of each index in the energy condition, water vapor condition, and dynamic condition are determined by normalization.

[0007] According to the present invention, a method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms is provided, wherein the selection of the pressure layer temperature difference index as the core index of energy conditions includes: The temperature difference between the first and second pressure layers is calculated, and combined with the dew point temperature of the first and third pressure layers, a temperature difference index reflecting the thermal properties of the atmosphere is obtained. The selection of lower-level water vapor flux as the core indicator of water vapor conditions includes: The flux value, which characterizes the mass of water vapor passing through a unit area per unit time, is calculated based on air density, specific humidity, and horizontal wind speed.

[0008] According to the present invention, a method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms includes calculating a comprehensive similarity index between the target region and each historical rainstorm region based on the background field similarity index system, comprising: Calculate the individual similarity coefficients between the target area and each historical rainstorm area under the energy conditions, water vapor conditions, and dynamic conditions respectively; The overall similarity index is obtained by weighting and summing the individual similarity coefficients based on the weight coefficients of each indicator.

[0009] According to the present invention, a method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms includes, in which the meteorological parameters at each level after coordinate transformation are physically constrained and corrected according to the topographic features of the target area, the method comprises: Perform water vapor balance constraint correction and adjust the distribution of lower-level water vapor flux according to the topographic slope aspect of the target area; Perform energy conservation constraint correction and adjust the atmospheric temperature field so that the deviation between the total energy of the atmospheric column in the adjusted target area and the energy of the similar background field of the transplanted object is within the allowable deviation range of energy conservation. Perform dynamic coordination constraint corrections to adjust wind speed and direction between different pressure layers, so that vertical wind shear is kept within the dynamic coordination shear threshold. The implementation of water vapor balance constraint correction, which adjusts the distribution of lower-level water vapor flux according to the topographic slope aspect of the target area, includes: Identify the windward and leeward slopes in the terrain of the target area; Increase the water vapor flux value at the corresponding location in the windward slope area, and the increase shall not exceed the upper limit ratio of the windward slope humidification. Reduce the water vapor flux at the corresponding location in the leeward slope area, and the reduction shall not exceed the upper limit of the leeward slope dehumidification ratio.

[0010] According to the present invention, a method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms is provided. The method involves performing coordinate transformation on the meteorological parameters at each level of the transplanted object, translating the latitude and longitude coordinates of the historical rainstorm area where the transplanted object is located to the center of the target area, and rotating and adjusting according to the topographic orientation and circulation characteristics of the target area to obtain the meteorological parameters at each level after coordinate transformation. This includes: The Lambert projection coordinate system was used for transformation, and the translation difference between the latitude and longitude of the target area center and the latitude and longitude of the historical rainstorm area center was calculated. The translation difference is superimposed onto the original latitude and longitude data of the transplanted object to obtain the translated latitude and longitude. The rotation center is determined as the center of the target area. The rotation angle is determined according to the terrain orientation and circulation characteristics of the target area. The translated latitude and longitude are calculated clockwise to obtain the coordinate transformation latitude and longitude, and the latitude and longitude deviation range is controlled.

[0011] According to the present invention, a method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms is provided, wherein the meteorological data of the transplanted object, after preprocessing, is decomposed into meteorological parameters at each level according to key pressure layers, including: The preprocessed meteorological data of the transplanted object is decomposed into multiple key pressure layers in the vertical direction, and the key pressure layers cover at least the pressure range from the near-surface layer to the upper atmosphere. For each key pressure layer, geopotential height, temperature, specific humidity, east-west wind speed component, and north-south wind speed component are extracted as meteorological parameters for each level.

[0012] According to the present invention, a method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms is provided. The method involves selecting suitable WRF models based on the underlying surface characteristics, climate background, and historical rainstorm causal types of the target area, including: Based on the types of causes of the historical rainstorms, suitable microphysical process schemes and cumulus convection parameterization schemes are selected from the alternative parameterization scheme library. Based on the underlying surface characteristics of the target area, a boundary layer scheme that is suitable for the boundary layer structure characteristics of complex terrain is selected from the candidate parameterized scheme library; Based on the climate background of the target area, long-wave and short-wave radiation schemes that are suitable for the atmospheric thermal background field are selected from the alternative parameterization scheme library. The selected microphysical process schemes, cumulus convection parameterization schemes, boundary layer schemes, and long and short wave radiation schemes are combined to construct the adapted WRF mode.

[0013] According to the physical mechanism-based method for porting extreme rainstorm spatiotemporal processes provided by the present invention, before running the WRF model, the method further includes: Determine the time span of historical rainstorm events in the target area; Based on the WRF model, the historical rainstorm event was simulated throughout the time span to obtain simulated precipitation data. Obtain measured precipitation data from rain gauges within the time span of the historical rainstorm event; Calculate the correlation coefficient, root mean square error, false alarm rate, and false alarm rate between the simulated precipitation data and the measured precipitation data; When the correlation coefficient, root mean square error, false alarm rate, and false alarm rate all meet the simulation verification qualification criteria, the WRF mode verification is confirmed to be successful.

[0014] This invention provides a physical mechanism-based method for transplanting spatiotemporal processes of extreme rainstorms. By selecting transplantation targets based on background field similarity and combining coordinate transformation and physical constraint correction to reconstruct meteorological parameters, it achieves the scientific transplantation of extreme rainstorm events. By introducing physical constraints such as water vapor balance, energy conservation, and dynamic coordination, it effectively solves the problems of missing physical processes and environmental mismatch caused by direct translation or transposition methods in traditional approaches, preventing the transplantation results from becoming a mere numbers game. The rainstorm process data generated by this method retains the disaster-causing intensity of historical extreme events and conforms to the actual topography and circulation characteristics of the target area, providing a high-value reference with physical authenticity for regional flood control planning and emergency drills. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0016] Figure 1 This is one of the flowcharts of the method for transplanting the spatiotemporal process of extreme rainstorms based on physical mechanisms provided by the present invention.

[0017] Figure 2 This is the second flowchart of the method for transplanting the spatiotemporal process of extreme rainstorms based on physical mechanisms provided by this invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0019] Figure 1 This is one of the flowcharts illustrating the method for transplanting the spatiotemporal process of extreme rainstorms based on physical mechanisms provided by this invention. Figure 2 This is the second flowchart illustrating the method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms provided by this invention. Figure 1 and Figure 2 As shown, the method includes steps 110, 120, 130, 140, 150 and 160.

[0020] Step 110: Based on the underlying surface characteristics, climate background, and historical rainfall patterns of the target area, select suitable WRF models.

[0021] Here, the target area refers to a specific geographical region requiring extreme rainstorm simulation and risk assessment, such as a specific watershed, city, or administrative region. Underlying surface characteristics are important physical attributes influencing local microclimate and precipitation distribution, specifically including topography (e.g., plains, mountains, hills), vegetation cover type (e.g., forests, grasslands, farmland), soil type, and moisture distribution. Climate background defines the large-scale circulation environment of the target area, such as whether the region is tropical or temperate, monsoon or non-monsoon. The causal type of historical rainstorms refers to the classification of the physical mechanisms causing extreme precipitation, such as convective rainstorms, frontal rainstorms, or typhoon rainstorms.

[0022] Selecting a suitable WRF model refers to choosing the combination that best reflects the meteorological characteristics and heavy rainfall events in the region from the numerous physical parameterization schemes available for the WRF (Weather Research and Forecasting) model, based on the aforementioned characteristics. The WRF model is a mesoscale numerical weather prediction system, and its simulation performance depends on the selection of the parameterization scheme.

[0023] For example, different boundary layer and radiation schemes need to be selected for different underlying surface features and climate backgrounds. If the target area has complex topography, schemes that optimize the simulation of boundary layer turbulent mixing, water vapor transport, and energy exchange processes, such as the YSU (Yonsei University) scheme, should be prioritized. For different causal types, such as torrential rain dominated by mesoscale convective systems, the selection of cumulus convection parameterization scheme is crucial, and the Kain-Fritsch scheme may be necessary; for large-scale frontal torrential rain, the Grell-3D scheme may be more suitable. The selection of microphysical process schemes directly affects the evolution of precipitation particles. Schemes such as WSM6 (WRF Single-Moment 6-class) should be selected according to the precipitation intensity and type to accurately characterize microphysical processes such as raindrop condensation, collision, and melting.

[0024] By selectively choosing suitable solutions, we can ensure that the WRF mode has a good physical foundation in subsequent simulations and avoid simulation deviations caused by improper mode configuration.

[0025] Step 120: Obtain meteorological data of historical rainstorms, construct a background field similarity index system that includes energy conditions, water vapor conditions and dynamic conditions, calculate the comprehensive similarity index between the target area and each historical rainstorm area based on the background field similarity index system, and take the historical rainstorm areas with a comprehensive similarity index greater than the similarity judgment threshold as the transplantation objects.

[0026] Specifically, the historical heavy rainfall meteorological data obtained usually comes from global or regional reanalysis datasets, such as NCEP FNL (Final Operational Global Analysis) global reanalysis data (e.g., spatial resolution 0.25°×0.25°, temporal resolution 6 hours) as the basic data, which provides high-resolution grid data containing multiple elements such as geopotential height, temperature, wind field, and humidity.

[0027] To ensure the transplanted rainstorm has a physical probability of occurring in the target area, this embodiment does not directly transplant any random rainstorm, but first performs a background field similarity analysis. The constructed background field similarity index system includes three core dimensions: Energy conditions: Reflecting the degree of atmospheric instability, they determine the potential for convection development within a watershed. Commonly used indicators include the K-index, which integrates temperature and dew point temperature information.

[0028] Water vapor conditions: directly constrain the intensity and duration of precipitation, and are the material basis for the formation of rainstorms. Commonly used indicators include water vapor flux in the lower atmosphere (e.g., 850 hPa).

[0029] Dynamic conditions: The lifting mechanism of the prevailing airflow and the configuration of weather systems influence the location and shape of the rainband. Commonly used indicators may include geopotential height in the middle layer (e.g., 500 hPa) and wind speed and direction in the lower layer.

[0030] The process of calculating the comprehensive similarity index based on the background field similarity index system is a quantitative evaluation process. First, the individual similarities between the current background field index of the target area and the background field index of other areas during historical extreme rainfall events are calculated. Weighting coefficients for each index can be determined through normalization, for example, assigning different weighting coefficients (e.g., 0.35, 0.35, 0.15, 0.15) to the K-index, water vapor flux, geopotential height, and wind speed and direction. Then, the comprehensive similarity index S is calculated using a weighted summation formula. The value of S typically ranges from 0 to 1; the closer the value is to 1, the more similar the meteorological backgrounds of the two regions.

[0031] The similarity threshold is a key criterion for determining whether transplantation is feasible. As an optional implementation, this similarity threshold can be set to 0.8. When the calculated comprehensive similarity index S ≥ 0.8, the target area is deemed to possess a similar environment to the historical rainstorm, and the historical rainstorm area and its corresponding rainstorm event are identified as the transplantation object. Conversely, if the similarity is too low, forced transplantation may violate physical laws, leading to unreliable simulation results.

[0032] Step 130: Based on the meteorological data of the preprocessed transplanted object, decompose the meteorological parameters of each level according to the key pressure layer.

[0033] After identifying the target data for transplantation, the raw meteorological data needs to be preprocessed and decomposed. Preprocessing may include format conversion, projection transformation (such as using Lambert-conformal projection with the center longitude aligned with the target area center) of the raw data using a WRF preprocessing system (i.e., WPS), horizontal interpolation (such as interpolating 0.25°×0.25° data to the target grid resolution, such as 3km / 9km / 27km), and temporal interpolation (generating hourly initial and boundary fields).

[0034] Decomposition by key pressure layers refers to the vertical stratification of the atmosphere. The atmosphere is not a homogeneous medium; meteorological elements vary greatly at different altitudes. In this embodiment, the meteorological data of the transplanted object is decomposed vertically into multiple key pressure layers, including 32 layers such as 50hPa, 100hPa, 200hPa, 500hPa, 700hPa, 850hPa, 925hPa, and the surface layer.

[0035] The meteorological parameters at each level, broken down into their respective layers, mainly include core variables characterizing atmospheric conditions, such as geopotential height, temperature, specific humidity, and horizontal wind speed. Horizontal wind speed is typically further decomposed into east-west and north-south wind speed components. In addition, surface layer parameters include surface air pressure, surface temperature, surface humidity, vegetation cover, and soil moisture.

[0036] Step 140: Perform coordinate transformation on the meteorological parameters of each level of the transplanted object, translate the latitude and longitude coordinates of the historical rainstorm area where the transplanted object is located to the center of the target area, and rotate and adjust according to the topographic orientation and circulation characteristics of the target area to obtain the meteorological parameters of each level after coordinate transformation.

[0037] Since the object being migrated is located in a different location, directly using its data would lead to positional errors. Therefore, coordinate transformation is required, which mainly involves two operations: translation and rotation.

[0038] Translation aims to move the center of a rainstorm to a target area. However, translation alone is insufficient. Different regions often have different topographic orientations and circulation characteristics. To ensure the transplanted rainstorm system adapts to the geographical environment of the target area, rotation adjustment is necessary. The rotation angle θ is determined based on the topographic orientation and circulation characteristics of the target area, typically ranging from 0° to 90°. Using the center of the target area as the origin of rotation, the translated coordinates are rotated clockwise or counterclockwise to obtain new latitude and longitude coordinates.

[0039] Step 150: Replace the original atmospheric boundary field parameters of the corresponding level in the target area with the meteorological parameters of each level after coordinate transformation, and perform physical constraint correction on the meteorological parameters of each level after coordinate transformation according to the terrain characteristics of the target area.

[0040] Specifically, after obtaining the meteorological parameters at each level after coordinate transformation, the meteorological parameters at each level are mapped and replaced with the original atmospheric initial field and boundary field data of the target area in the WRF model to ensure that the transplanted atmospheric boundary field is compatible with the underlying surface features (topography, vegetation, soil) of the target area.

[0041] However, simple substitution can lead to physical inconsistencies. For example, the original rainstorm area might be a plain, while the target area might be a mountainous region. The airflow and moisture distribution in the original data are adapted to plains; directly applying them to mountains would violate principles of fluid dynamics and thermodynamics. Therefore, physical constraint corrections are necessary.

[0042] Step 160: Use the meteorological parameters at each level after physical constraint correction as the initial field and boundary field input parameters of the WRF model, run the WRF model, and generate spatiotemporal process data of extreme rainstorms in the target area.

[0043] After completing all the above processing steps, a new meteorological dataset was obtained that not only includes historical extreme rainfall characteristics but also adapts to the topography and physical environment of the target area. This dataset was then used as the input parameters for the initial and boundary fields into the adapted WRF model constructed in step 110.

[0044] The WRF model is run for numerical integration calculations. Based on atmospheric dynamic equations, the model simulates the weather evolution of the target area over a future period under the current extreme conditions. The final output data represents the spatiotemporal process data of the extreme rainfall event in the target area, including hourly rainfall distribution, rainband movement path, and evolution of rainfall center intensity. Analyzing this data provides a clear view of the spatiotemporal distribution characteristics that would occur if this historically rare extreme rainfall event were to actually occur locally.

[0045] After generating the spatiotemporal data of extreme rainfall events in the target area, analysis can be performed from both temporal and spatial dimensions, as detailed below: Time dimension analysis: Extract hourly precipitation sequences of simulated rainstorm processes and rainfall characteristic values ​​(maximum basin rainfall, maximum hourly rainfall, etc.) of different transplantation schemes, and analyze the temporal evolution characteristics of rainstorms and compare them with the temporal characteristics of historical extreme rainstorms in the target area (such as average peak occurrence time and concentrated precipitation periods).

[0046] Spatial dimension analysis: Plot the hourly spatial distribution of simulated rainstorm precipitation, analyze the spatial location, intensity, and coverage of the rainstorm center, and verify whether the rain belt follows the topography and the distribution of water vapor channels.

[0047] This embodiment provides a method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms. By selecting transplantation targets based on background field similarity and combining coordinate transformation and physical constraint correction to reconstruct meteorological parameters, it achieves the scientific transplantation of extreme rainstorm events. By introducing physical constraints such as water vapor balance, energy conservation, and dynamic coordination, it effectively solves the problems of missing physical processes and environmental mismatch caused by direct translation or transposition methods in traditional approaches, avoiding the transplantation results becoming a mere numbers game. The rainstorm process data generated by this method not only retains the disaster-causing intensity of historical extreme events but also conforms to the actual topography and circulation characteristics of the target area, providing a high-value reference with physical authenticity for regional flood control planning and emergency drills.

[0048] Based on the above embodiments, a background field similarity index system including energy conditions, water vapor conditions, and dynamic conditions is constructed, including: The pressure layer temperature difference index was selected as the core index of energy conditions. Lower-level water vapor flux was selected as the core indicator of water vapor conditions. High-level geopotential height and low-level wind speed and direction are selected as the core indicators of dynamic conditions. The weighting coefficients of each index in the energy condition, water vapor condition, and dynamic condition are determined by normalization. Among them, the pressure layer temperature difference index is selected as the core index of energy conditions, including: The temperature difference between the first and second pressure layers is calculated, and combined with the dew point temperature of the first and third pressure layers, a temperature difference index reflecting the thermal properties of the atmosphere is obtained. Lower-level water vapor flux was selected as the core indicator of water vapor conditions, including: The flux value, which characterizes the mass of water vapor passing through a unit area per unit time, is calculated based on air density, specific humidity, and horizontal wind speed.

[0049] The reason for using the barostratus temperature difference index is that atmospheric stratification instability represents the energy potential for heavy rainfall. A greater temperature difference between upper and lower layers leads to a more significant difference in air buoyancy, which is more conducive to the outbreak of severe convective weather. Therefore, by selecting the barostratus temperature difference index, the degree of thermal instability in the vertical direction of the atmosphere can be effectively quantified.

[0050] Here, the barosphere temperature difference index refers to a parameter that characterizes the static stability of the atmosphere by using the temperature difference between different pressure layers at different altitudes. As an optional embodiment, this embodiment selects the K-index as the specific implementation form of the barosphere temperature difference index. The K-index comprehensively considers the temperature lapse rate of the middle and lower atmosphere and the humidity conditions of the lower atmosphere, and its calculation formula is as follows: K=T850-T500+Td850-(T700-Td700); Wherein, T850, T500, and T700 are the temperatures of the first pressure layer at 850 hPa, the second pressure layer at 500 hPa, and the third pressure layer at 700 hPa, respectively; and Td850 and Td700 are the dew point temperatures of the first pressure layer at 850 hPa and the third pressure layer at 700 hPa, respectively. Thresholds are set according to the climate type of the target area: K ≥ 35℃ in tropical monsoon regions indicates a highly unstable state, and K ≥ 30℃ in temperate regions indicates a highly unstable state.

[0051] After determining the evaluation indicators for energy conditions, considering that energy potential alone is insufficient to form rainstorms, and that abundant water vapor transport is the material basis for extreme precipitation, this embodiment further selects lower-level water vapor flux as the core indicator of water vapor conditions.

[0052] Here, lower-level water vapor flux refers to the mass of water vapor passing through a unit area per unit time. As an optional embodiment, this embodiment selects the 850 hPa water vapor flux as the core indicator, and its calculation formula is as follows: Q = (1 / ρ) × q × v; Where ρ is air density, q is specific humidity, and v is horizontal wind speed. A threshold is set: a water vapor flux ≥ 10 g / (cm·hPa·s) is considered a state of abundant water vapor, meeting the water vapor requirements for the formation of extreme rainstorms.

[0053] After clarifying the energy and water vapor conditions, and considering that the location and form of the rainstorm are also constrained by the atmospheric circulation pattern, in order to ensure that the location of the transplanted rain belt and the configuration of the weather system are reasonable, this embodiment further selects the upper geopotential height and the lower wind speed and direction as the core indicators of the dynamic conditions.

[0054] As an optional embodiment, this embodiment selects 500hPa geopotential height and 850hPa wind speed and direction as core indicators.

[0055] For the 500hPa geopotential height, the geopotential height values ​​of grid points are used. The difference in geopotential height (ΔH) between the target area and the historical rainstorm area is calculated. ΔH ≤ 50gpm indicates a good match. For wind speed and direction at 850 hPa, calculate the wind direction deviation angle (Δθ) and the relative wind speed deviation (ΔV / V), where Δθ≤30° and ΔV / V≤20% indicate a good match.

[0056] After establishing the above three categories of core indicators, considering the huge differences in the dimensions and numerical ranges of each indicator, it is impossible to make a direct comprehensive comparison. Therefore, this embodiment determines the weight coefficients of each indicator in energy conditions, water vapor conditions and dynamic conditions through normalization processing.

[0057] As an optional embodiment, after normalization, the weight coefficients of each indicator are set as follows: the weight coefficient of K index is 0.35, the weight coefficient of 850hPa water vapor flux is 0.35, the weight coefficient of 500hPa geopotential height is 0.15, and the weight coefficient of 850hPa wind speed and direction is 0.15.

[0058] Based on the above embodiments, the comprehensive similarity index between the target area and each historical rainstorm area is calculated based on the background field similarity index system, including: Calculate the individual similarity coefficients between the target area and each historical rainstorm area in terms of energy conditions, water vapor conditions, and dynamic conditions; Based on the weight coefficients of each indicator, the individual similarity coefficients are weighted and summed to obtain the comprehensive similarity index.

[0059] Here, the single-item similarity coefficient refers to the value that characterizes the similarity between a certain meteorological index value in the target area and the corresponding index value in the historical rainstorm area.

[0060] As an optional embodiment, the individual similarity coefficient can be calculated using the following formula: ; in, For the first The individual similarity coefficient of each indicator. For the target area in the first The measured values ​​for each indicator For historical rainstorm areas during the historical rainstorm period The value of each indicator, This is the historical maximum value of this indicator.

[0061] After obtaining the individual similarity coefficients between the target area and each historical rainstorm area under various conditions, considering that different physical conditions contribute differently to the formation of rainstorms, in order to obtain a final evaluation result that can represent the overall environmental similarity, this embodiment further performs a weighted summation of the individual similarity coefficients based on the weight coefficients of each indicator to obtain a comprehensive similarity index.

[0062] The comprehensive similarity index can be calculated based on the following formula: ; in, This represents the comprehensive similarity index. For the first The weighting coefficients of each indicator, when At a value of 0.8, it is determined to be a similar background field, which can be used for subsequent rainstorm transplantation simulation.

[0063] Based on the above embodiments, physical constraint corrections are performed on the meteorological parameters at each level after coordinate transformation according to the terrain features of the target area, including: Perform water vapor balance constraint correction and adjust the distribution of lower-level water vapor flux according to the topographic slope aspect of the target area; Perform energy conservation constraint correction and adjust the atmospheric temperature field so that the deviation between the total energy of the atmospheric column in the adjusted target area and the energy of the similar background field of the transplanted object is within the allowable deviation range of energy conservation. Perform dynamic coordination constraint corrections to adjust wind speed and direction between different pressure layers, so that vertical wind shear remains within the dynamic coordination shear threshold.

[0064] Among these, adjusting the distribution of lower-level water vapor flux according to the topographic slope aspect of the target area includes: Identify the windward and leeward slopes in the target area's terrain.

[0065] Specifically, physical constraint correction mainly includes the following aspects: Energy conservation constraint: When adjusting temperature field data, the law of conservation of energy must be followed. By calculating the total energy of the atmospheric column before and after adjustment, ensure that the deviation between the total energy of the atmospheric column in the target area after adjustment and the energy of the similar background field of the transplanted object is within the allowable deviation range of energy conservation (e.g., ≤5%).

[0066] Water vapor balance constraint: Based on the topographic features of the target area, especially the topographic slope (such as windward slope and leeward slope), adjust the water vapor flux distribution at 850 hPa. The increase in water vapor flux on the windward slope shall not exceed 30% of the upper limit of the windward slope humidification, and the decrease on the leeward slope shall not exceed 20% of the upper limit of the leeward slope dehumidification, so as to ensure that the water vapor distribution matches the topography.

[0067] Dynamic coordination constraints: Optimize the vertical shear of wind speed and direction at 500hPa and 850hPa to ensure that the shear value is within the shear range of historical extreme rainstorms in the target area (shear value deviation ≤10%), and maintain the coordination of the dynamic field.

[0068] Based on the above embodiments, coordinate transformation is performed on the meteorological parameters at each level of the transplanted object. The latitude and longitude coordinates of the historical rainstorm area where the transplanted object is located are translated to the center of the target area, and rotation adjustment is performed according to the topographic orientation and circulation characteristics of the target area to obtain the meteorological parameters at each level after coordinate transformation, including: The Lambert projection coordinate system was used for transformation, and the translation difference between the latitude and longitude of the target area center and the latitude and longitude of the historical rainstorm area center was calculated. The translation difference is superimposed onto the original latitude and longitude data of the transplanted object to obtain the translated latitude and longitude. The rotation center is determined as the center of the target area. The rotation angle is determined according to the terrain orientation and circulation characteristics of the target area. The latitude and longitude after translation are calculated clockwise to obtain the latitude and longitude after coordinate transformation, and the range of latitude and longitude deviation is controlled.

[0069] Specifically, the translation operation aims to move the center of the rainstorm to the target area. Specifically, the difference (dx, dy) between the center latitude and longitude of the target area (Lon_target, Lat_target) and the center latitude and longitude of the historical rainstorm area (Lon_hist, Lat_hist) can be calculated, i.e., dx = Lon_target - Lon_hist, dy = Lat_target - Lat_hist. Then, this difference is added to the latitude and longitude of all grid points in the original data of the transplanted object, thereby achieving the overall translation of the rainstorm system. That is, the transformed coordinates are Lon_new = Lon_old + dx, Lat_new = Lat_old + dy, where Lon_old and Lat_old are the original data latitude and longitude, and Lon_new and Lat_new are the translated latitude and longitude.

[0070] However, translation alone is insufficient. Different regions often exhibit varying topographic orientations and circulation characteristics. To ensure the transplanted rainstorm system adapts to the target region's geographical environment, rotation adjustment is necessary. The rotation angle θ is determined based on the topographic orientation and circulation characteristics of the target region, typically ranging from 0° to 90°. Using the center of the target region as the origin of rotation, the translated coordinates are rotated clockwise or counterclockwise to obtain new latitude and longitude coordinates. The rotated coordinates can be expressed as: Lon_rot=(Lon_new-Lon_target)×cosθ+(Lat_new-Lat_target)×sinθ+Lon_target; Lat_rot=-(Lon_new-Lon_target)×sinθ+(Lat_new-Lat_target)×cosθ+Lat_target; As a preferred implementation, when performing rotational adjustments, it is also necessary to control the deviation between the latitude and longitude range after transplantation and the target area. For example, the maximum latitude and longitude deviation should be controlled to not exceed 10°, i.e., |Lon_rot-Lon_target|≤10°, |Lat_rot-Lat_target|≤10°, to avoid the circulation conditions deviating from the climate background of the target area due to the transplantation range being too large.

[0071] Based on the above embodiments, based on the preprocessed meteorological data of the transplanted object, meteorological parameters at each level are decomposed according to the key pressure layer, including: The preprocessed meteorological data of the transplanted object is decomposed into multiple key pressure layers in the vertical direction. The key pressure layers cover at least the pressure range from the near-surface layer to the upper atmosphere. For each key pressure layer, geopotential height, temperature, specific humidity, east-west wind speed component, and north-south wind speed component are extracted as meteorological parameters for each level.

[0072] Numerical weather prediction models (such as WRF) are based on three-dimensional grids and require initial states at different altitudes as input. Covering the entire pressure layer from near-surface to upper atmosphere ensures the integrity of atmospheric column information and avoids model instability due to missing information. Therefore, this embodiment decomposes the preprocessed meteorological data of the transplanted object into multiple key pressure layers in the vertical direction.

[0073] Here, the key pressure layer refers to a meteorologically representative isobaric surface, which is usually used to characterize the atmospheric circulation patterns at different altitudes.

[0074] As an optional embodiment, this embodiment breaks down meteorological data into 32 layers. These key pressure layers specifically include, but are not limited to: 50 hPa, 100 hPa, 200 hPa, 500 hPa, 700 hPa, 850 hPa, 925 hPa, and the surface layer. This layered setup covers the entire troposphere from the surface to the bottom of the stratosphere, ensuring a full three-dimensional structural description of the rainfall system.

[0075] After completing the vertical layering, considering that the atmospheric state of each layer needs to be described by specific physical variables for numerical model recognition and calculation, this embodiment further extracts geopotential height, temperature, specific humidity, east-west wind speed component, and north-south wind speed component as meteorological parameters for each key pressure layer. These five parameters constitute a complete set describing the dynamic and thermal state of the atmosphere. Geopotential height reflects the pressure field distribution, temperature reflects the thermal structure, specific humidity reflects the water vapor content, and wind speed component reflects dynamic transmission.

[0076] Based on the above embodiments, and considering the underlying surface characteristics, climate background, and historical rainfall patterns of the target area, suitable WRF models are selected, including: Based on the types of historical rainstorms, suitable microphysical process schemes and cumulus convection parameterization schemes were selected from the alternative parameterization scheme library. Based on the underlying surface characteristics of the target area, boundary layer schemes that are suitable for the complex terrain boundary layer structure characteristics are selected from the alternative parametric scheme library; Based on the climate background of the target area, long-wave and short-wave radiation schemes that are suitable for the atmospheric thermal background field are selected from the alternative parameterization scheme library. The selected microphysical process schemes, cumulus convection parameterization schemes, boundary layer schemes, and long and short wave radiation schemes are combined to construct an adapted WRF model.

[0077] Here, the microphysics process scheme is used to calculate the interconversion of particles such as water vapor, cloud water, rain, snow, and graupel in the atmosphere; the cumulus convection parameterization scheme is used to handle subgrid convection processes below the grid point scale.

[0078] As an optional embodiment, this embodiment, based on the specific characteristics of the target scenario, uses simulation to select a combination of schemes that accurately reflect the physical processes of meteorological conditions. Specifically, for different types of historical rainstorms (such as mesoscale convective system-dominated or large-scale frontal rainstorms), a cumulus convection parameterization scheme that can balance the accuracy of convection triggering and vertical motion simulation, and a microphysics scheme that can accurately characterize complex microphysical processes are selected; for the underlying surface characteristics of the target area (such as complex terrain), a boundary layer scheme that can optimize the simulation of boundary layer energy exchange is selected; for the climate background of the target area, a long-wave and short-wave radiation scheme that is adapted to the atmospheric thermal background field is selected. Finally, the selected sub-schemes are combined to construct an adapted WRF model to ensure the output of optimal simulation results under different transplantation scenarios.

[0079] Considering that a single-resolution grid cannot simultaneously capture both the breadth of the simulation area and the accuracy of the core region, this embodiment further employs a multi-layer nested simulation technique when constructing the adapted WRF mode to ensure computational efficiency while accurately capturing the small-to-medium-scale characteristics of extreme rainstorms in the target area. Here, multi-layer nesting refers to setting up multiple simulation regions with progressively increasing resolutions within the WRF mode.

[0080] As an optional embodiment, this embodiment sets up three nested regions, namely regions D01, D02 and D03.

[0081] D01 (outermost region): Covers the area including the target area and a large surrounding environmental field. It adopts a coarser spatial resolution to simulate the evolution of large-scale weather systems and provide background field forcing.

[0082] D02 (Intermediate Layer Region): Nested within D01, it covers the target watershed and surrounding key influence areas. It uses medium resolution to convey large-scale information and perform preliminary analysis of the mesoscale system.

[0083] D03 (innermost region): Nested inside D02, it precisely covers the target area and uses high resolution to finely simulate the location, intensity, and interaction with complex terrain of extreme rainstorms.

[0084] Based on the above embodiments, before running WRF mode, the following is also included: Determine the time span of historical rainstorm events in the target area; Based on the WRF model, historical rainstorm events were simulated over a time span to obtain simulated precipitation data. Obtain measured precipitation data from rain gauges over a historical time span of heavy rain events; Calculate the correlation coefficient, root mean square error, false alarm rate, and false alarm rate between simulated precipitation data and measured precipitation data; When the correlation coefficient, root mean square error, false alarm rate, and false alarm rate all meet the simulation verification qualification criteria, the WRF mode verification is confirmed to be successful.

[0085] Here, historical rainstorm events refer to extreme precipitation events that have actually occurred in the target area in the past. The time span refers to the complete period from the beginning to the end of the impact of the event, which includes the entire stage of rainstorm incubation, development, peak, and decline.

[0086] As an optional implementation, typical historical rainstorm events with the largest rainfall in the target area over the past 10 years can be screened to determine their time span. For example, if a rainstorm occurred between July 20 and July 22, the time span is set to 20:00 on July 19 to 08:00 on July 23 to ensure that the entire process of the rainstorm's incubation, development, peak, and decline is included.

[0087] After determining the simulation object and time range, this embodiment further simulates historical rainstorm events over the entire time span based on the WRF model to obtain simulated precipitation data. Here, simulated precipitation data refers to the hourly gridded rainfall data output by the WRF model.

[0088] As an optional embodiment, the adapted WRF model constructed above is used, and the original meteorological reanalysis data of the historical rainstorm event is input as the initial field and boundary field. The model simulation domain, vertical layer number, and simulation duration are set. Numerical integration is performed within the set time span, and the hourly simulated precipitation data of each grid point in the target area is output.

[0089] In addition to obtaining the simulation results, this embodiment further acquires historical rainstorm event data from rain gauges over a given time span to provide a standard comparison. Here, the rain gauge data refers to the hourly rainfall observed by high-density rain gauges within the target area.

[0090] Having obtained both the simulated data and the measured data for the standard answer, and considering that a single indicator is insufficient to fully reflect the simulation bias, this embodiment further calculates the correlation coefficient (CC), root mean square error (RMSE), weather TS score, false alarm rate (FAR), and false alarm rate between the simulated precipitation data and the measured precipitation data, focusing on verifying the consistency between the rainfall amount and the spatiotemporal distribution of precipitation and the actual observations.

[0091] After calculating the various verification indicators, this embodiment further determines that the WRF model verification is successful when the correlation coefficient, root mean square error, false alarm rate, and false alarm rate all meet the simulation verification pass criteria. It is understood that if any indicator fails to meet the simulation verification pass criteria, the WRF model's parameterization scheme combination is adjusted, such as by changing the cumulus convection parameterization scheme, and the simulation verification is performed again.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for transplanting the spatiotemporal process of extreme rainstorms based on physical mechanisms, characterized in that, include: Based on the underlying surface characteristics, climate background, and historical rainfall patterns of the target area, suitable weather numerical simulation (WRF) models are selected. Meteorological data of historical rainstorms are acquired, and a background field similarity index system including energy conditions, water vapor conditions and dynamic conditions is constructed. Based on the background field similarity index system, the comprehensive similarity index between the target area and each historical rainstorm area is calculated, and historical rainstorm areas with a comprehensive similarity index greater than the similarity judgment threshold are used as transplantation objects. Based on the preprocessed meteorological data of the transplanted object, meteorological parameters of each level are decomposed according to the key pressure layer. The meteorological parameters at each level of the transplanted object are transformed by coordinate transformation. The latitude and longitude coordinates of the historical rainstorm area where the transplanted object is located are translated to the center of the target area. The rotation is adjusted according to the topography and circulation characteristics of the target area to obtain the meteorological parameters at each level after coordinate transformation. The meteorological parameters at each level after coordinate transformation are used to replace the original atmospheric boundary field parameters at the corresponding level in the target area, and the meteorological parameters at each level after coordinate transformation are physically constrained and corrected according to the topographic features of the target area. The meteorological parameters at each level, after being corrected by physical constraints, are used as the initial field and boundary field input parameters of the WRF model. The WRF model is then run to generate spatiotemporal process data of extreme rainstorms in the target area. The step of physically constraining and correcting the meteorological parameters at each level after coordinate transformation based on the terrain features of the target area includes: Perform water vapor balance constraint correction and adjust the distribution of lower-level water vapor flux according to the topographic slope aspect of the target area; Perform energy conservation constraint correction and adjust the atmospheric temperature field so that the deviation between the total energy of the atmospheric column in the adjusted target area and the energy of the similar background field of the transplanted object is within the allowable deviation range of energy conservation. Perform dynamic coordination constraint corrections to adjust wind speed and direction between different pressure layers, so that vertical wind shear is kept within the dynamic coordination shear threshold. The implementation of water vapor balance constraint correction, which adjusts the distribution of lower-level water vapor flux according to the topographic slope aspect of the target area, includes: Identify the windward and leeward slopes in the terrain of the target area; Increase the water vapor flux value at the corresponding location in the windward slope area, and the increase shall not exceed the upper limit ratio of the windward slope humidification. Reduce the water vapor flux at the corresponding location in the leeward slope area, and the reduction shall not exceed the upper limit of the leeward slope dehumidification ratio.

2. The method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms according to claim 1, characterized in that, The construction of a background field similarity index system, which includes energy conditions, water vapor conditions, and dynamic conditions, includes: The pressure layer temperature difference index was selected as the core index of energy conditions. Lower-level water vapor flux was selected as the core indicator of water vapor conditions. High-level geopotential height and low-level wind speed and direction are selected as the core indicators of dynamic conditions. The weighting coefficients of each index in the energy condition, water vapor condition, and dynamic condition are determined by normalization.

3. The method for transplanting the spatiotemporal process of extreme rainstorms based on physical mechanisms according to claim 2, characterized in that, The selection of atmospheric pressure layer temperature difference as the core indicator of energy conditions includes: The temperature difference between the first and second pressure layers is calculated, and combined with the dew point temperature of the first and third pressure layers, a temperature difference index reflecting the thermal properties of the atmosphere is obtained. The selection of lower-level water vapor flux as the core indicator of water vapor conditions includes: The water vapor flux per unit time per unit area is calculated based on air density, specific humidity, and horizontal wind speed.

4. The method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms according to claim 2, characterized in that, The calculation of the comprehensive similarity index between the target area and each historical rainstorm area based on the background field similarity index system includes: Calculate the individual similarity coefficients between the target area and each historical rainstorm area under the energy conditions, water vapor conditions, and dynamic conditions respectively; The overall similarity index is obtained by weighting and summing the individual similarity coefficients based on the weight coefficients of each indicator.

5. The method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms according to any one of claims 1 to 4, characterized in that, The process involves performing coordinate transformation on the meteorological parameters at each level of the transplanted object, shifting the latitude and longitude coordinates of the historical rainstorm area where the transplanted object is located to the center of the target area, and then rotating and adjusting them according to the topography and circulation characteristics of the target area to obtain the meteorological parameters at each level after coordinate transformation, including: The Lambert projection coordinate system was used for transformation, and the translation difference between the latitude and longitude of the target area center and the latitude and longitude of the historical rainstorm area center was calculated. The translation difference is superimposed onto the original latitude and longitude data of the transplanted object to obtain the translated latitude and longitude. The rotation center is determined as the center of the target area. The rotation angle is determined according to the terrain orientation and circulation characteristics of the target area. The translated latitude and longitude are calculated clockwise to obtain the coordinate transformation latitude and longitude, and the latitude and longitude deviation range is controlled.

6. The method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms according to any one of claims 1 to 4, characterized in that, The meteorological data of the transplanted object, after preprocessing, is decomposed into meteorological parameters at each level according to key pressure layers, including: The preprocessed meteorological data of the transplanted object is decomposed into multiple key pressure layers in the vertical direction, and the key pressure layers cover at least the pressure range from the near-surface layer to the upper atmosphere. For each key pressure layer, geopotential height, temperature, specific humidity, east-west wind speed component, and north-south wind speed component are extracted as meteorological parameters for each level.

7. The method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms according to any one of claims 1 to 4, characterized in that, The selection of suitable WRF models based on the underlying surface characteristics, climate background, and historical rainfall patterns of the target area includes: Based on the types of causes of the historical rainstorms, suitable microphysical process schemes and cumulus convection parameterization schemes are selected from the alternative parameterization scheme library. Based on the underlying surface characteristics of the target area, a boundary layer scheme that is suitable for the boundary layer structure characteristics of complex terrain is selected from the candidate parameterized scheme library; Based on the climate background of the target area, long-wave and short-wave radiation schemes that are suitable for the atmospheric thermal background field are selected from the alternative parameterization scheme library. The selected microphysical process schemes, cumulus convection parameterization schemes, boundary layer schemes, and long and short wave radiation schemes are combined to construct the adapted WRF mode.

8. The method for transplanting spatiotemporal processes of extreme rainstorms based on physical mechanisms according to any one of claims 1 to 4, characterized in that, Before running the WRF mode, the following is also included: Determine the time span of historical rainstorm events in the target area; Based on the WRF model, the historical rainstorm event was simulated throughout the time span to obtain simulated precipitation data. Obtain measured precipitation data from rain gauges within the time span of the historical rainstorm event; Calculate the correlation coefficient, root mean square error, false alarm rate, and false alarm rate between the simulated precipitation data and the measured precipitation data; When the correlation coefficient, root mean square error, false alarm rate, and false alarm rate all meet the simulation verification qualification criteria, the WRF mode verification is confirmed to be successful.

Images