WRF-based high-temporal-spatial-resolution drought and flood conversion prediction and early warning method and system

Abstract

The invention relates to a WRF-based high-temporal-spatial-resolution drought and flood conversion prediction and early warning method and system, and solves the problem of early warning lag caused by difficulty in accurately capturing subtle changes of drought and flood quick conversion, and the method comprises the following steps: a WRF model gradually advances calculation according to a model parameter setting scheme and an initial condition and a time step length to obtain a WRF model; simulating the movement and change process of the atmosphere; after the operation of the WRF model is finished, outputting prediction results of various meteorological elements, and evaluating and determining a drought and flood risk level according to the output prediction results of various meteorological elements and a preset drought and flood risk level analysis algorithm; and according to the corresponding relationship between the drought and flood risk level and the intelligent household equipment linkage scheme, determining the intelligent household equipment linkage scheme, and executing the intelligent household equipment linkage scheme. The method has the following effects: the drought and flood prediction precision is effectively improved, early warning is performed in time, the home equipment is linked, and the drought and flood disaster coping capability is improved.

Description

Technical Field

[0001] The present invention relates to the field of meteorology and hydrology intersection technology, and in particular to a WRF-based high temporal and spatial resolution drought-flood conversion prediction and early warning method and system. Background Art

[0002] High-temporal-resolution drought-flood conversion forecast and warning can provide decision makers with more accurate information so that they can rationally arrange resources, formulate emergency plans and take protective measures. A high-temporal-resolution drought-flood conversion forecast and warning method based on WRF (Weather Research and Forecasting) came into being. It comprehensively utilizes multiple data sources and advanced numerical simulation technology to improve the accuracy and timeliness of drought and flood prediction.

[0003] In existing technologies, drought and flood prediction mainly relies on traditional meteorological observation data and statistical analysis methods. Meteorological data such as temperature, precipitation, and air pressure are collected through meteorological observation stations, and statistical analysis is performed in combination with historical data to predict the possibility of drought and flood disasters. In addition, some studies have also used numerical weather forecast models, but these models are usually limited in temporal and spatial resolution and it is difficult to accurately capture the detailed changes in drought and flood conversion.

[0004] The key drawbacks of existing drought and flood prediction technology are as follows: First, the data source is single, mainly relying on data from meteorological observation stations, lacking consideration of multiple factors such as geographic information, and insufficient accuracy. Second, the low spatiotemporal resolution makes it difficult to accurately capture the subtle changes in the rapid conversion of drought and flood, resulting in delayed warnings. This makes it impossible to take timely and effective measures in response to drought and flood disasters, reduces the ability to prevent and reduce disasters, brings greater risks to people's lives and social and economic development, and cannot meet the current demand for high spatiotemporal resolution drought and flood conversion prediction and warning. Summary of the invention

[0005] In order to effectively improve the accuracy of drought and flood prediction, timely warn and link household appliances, and enhance the ability to cope with drought and flood disasters, the present application provides a high temporal and spatial resolution drought and flood conversion prediction and warning method and system based on WRF.

[0006] In the first aspect, the present application provides a high temporal and spatial resolution drought-flood conversion prediction and early warning method based on WRF, which adopts the following technical solutions:

[0007] A high temporal and spatial resolution drought and flood conversion prediction and early warning method based on WRF, including:

[0008] Obtain meteorological data, geographic information data and other data, and perform data preprocessing;

[0009] Obtain relevant parameter information of the area to be studied, and match and analyze the area category of the area to be studied based on the parameter information associated with different categories of areas;

[0010] According to the correspondence between the regional category and the model parameter setting scheme, the model parameter setting scheme of the area to be studied is analyzed and determined, and the model parameter setting scheme of the area to be studied is implemented;

[0011] The data assimilation algorithm is used to fuse data from different sources to generate initial field data. The boundary condition type of the area to be studied is matched and obtained according to the correspondence between the condition information of the area to be studied and the boundary condition type, and the boundary layer is processed using the boundary layer parameterization scheme.

[0012] The WRF model simulates the movement and change of the atmosphere by gradually advancing the calculation according to the model parameter setting scheme and initial conditions, where the initial conditions include the initial field setting and the boundary condition setting.

[0013] After the WRF model runs, it outputs the forecast results of various meteorological elements, and evaluates and determines the drought and flood risk level based on the forecast results of various meteorological elements and the preset drought and flood risk level analysis algorithm;

[0014] According to the correspondence between the drought and flood risk level and the smart home device linkage plan, the smart home device linkage plan is confirmed and implemented.

[0015] By adopting the above technical solution, we can obtain multi-source data and pre-process it, determine the regional category and model parameters, fuse the data to generate the initial field, use the WRF model to simulate atmospheric changes and output the forecast results of meteorological elements, and then link the smart home devices after evaluating the drought and flood risk level. This method can effectively improve the accuracy of drought and flood prediction, timely warn and link home devices, and enhance the ability to deal with drought and flood disasters.

[0016] Optionally, according to the correspondence between the region category and the model parameter setting scheme, analyzing and determining the model parameter setting scheme for the area to be studied, and executing the model parameter setting scheme for the area to be studied includes:

[0017] Analyze whether there are extreme meteorological events in the study area within the preset time period;

[0018] If yes, then obtain the data of historical extreme meteorological events, classify the extreme meteorological events according to the data of historical extreme meteorological events, and extract the characteristic parameters of the events;

[0019] Input the feature parameters of the extracted events into the trained model parameter setting matching model, and output the model parameter setting scheme;

[0020] If not, keep the original setting.

[0021] By adopting the above technical solution, when there are extreme meteorological events, by analyzing historical data for classification and extracting characteristic parameters, and using model parameter settings to match the model to determine the solution, the model can respond to extreme situations more accurately and improve the accuracy of drought and flood prediction. If there are no extreme events, the original settings are maintained to avoid unnecessary adjustments and save computing resources.

[0022] Optionally, according to the correspondence between the region category and the model parameter setting scheme, analyzing and determining the model parameter setting scheme for the area to be studied, and executing the model parameter setting scheme for the area to be studied includes:

[0023] Obtain change information of the area to be studied, including comprehensive land use change value, comprehensive infrastructure construction progress value, and ecological environment change value;

[0024] Analyze whether various values ​​of the change information exceed the preset change threshold of the corresponding category;

[0025] If yes, the regional variation characteristics are used as input objects, input into the relationship model between the trained physical process parameterization scheme and the regional variation characteristics, and output the adjustment parameters of the physical process parameterization scheme;

[0026] If not, the model parameter setting scheme for the area to be studied is analyzed and determined according to the correspondence between the region category and the model parameter setting scheme, and the model parameter setting scheme for the area to be studied is executed.

[0027] By adopting the above technical solution, by monitoring the change information of the area to be studied, changes in land use, infrastructure construction and ecological environment can be detected in time. If the change correlation value exceeds the threshold, the physical process parameterization scheme is adjusted using the relationship model, so that the model can better adapt to regional changes and improve the accuracy of drought and flood prediction. If it does not exceed the threshold, the model parameter setting scheme is determined in a conventional way to ensure the stability and efficiency of the method. It helps to more accurately predict and warn about drought and flood conversion, provide a reliable basis for regional development and the linkage of smart home devices, and enhance the ability to cope with drought and flood disasters.

[0028] Optionally, if the change information is a land use change value, obtaining the change information of the area to be studied includes:

[0029] Use satellite remote sensing images to obtain land use data of different expected study areas, and use image classification algorithms to divide the land into different use types, including cultivated land, forest land, construction land, and water areas;

[0030] Compare and analyze the land use classification results in different periods, and calculate the change area ratio and change speed of each land use type;

[0031] The two indicators of change area ratio and change speed are standardized, and according to the importance weights of different indicators, the standardized change area ratio and change speed are weighted summed to obtain the comprehensive change value of each land use type;

[0032] The comprehensive change values ​​of all land use types are summarized and summed to obtain the comprehensive value of land use change.

[0033] By adopting the above technical solution and using satellite remote sensing images to obtain land use data, we can accurately grasp the land use changes in the area to be studied. By classifying and comparing, calculating the proportion and speed of the changing area, and then performing standardization and weighted summation, we can obtain the comprehensive change value, and then obtain the comprehensive value of land use change. This helps to understand the regional land use dynamics more finely, provide an accurate basis for model parameter setting, and make drought and flood prediction and warning more targeted and accurate.

[0034] Optionally, if the change information is the progress value of comprehensive infrastructure construction, obtaining the change information of the area to be studied includes:

[0035] Collect and obtain relevant data on infrastructure construction in the area to be studied, including the length of road construction, building area, and number of public facilities;

[0036] According to the corresponding relationship between different infrastructure construction categories and key indicators, determine the key indicators and calculate the progress value of infrastructure construction;

[0037] The progress value of each infrastructure construction is standardized, and the standardized single progress value is indexed to obtain the single progress index;

[0038] The weighted sum of each individual progress index is used to obtain the comprehensive infrastructure construction progress index, and the comprehensive infrastructure construction progress value is obtained through index inverse transformation.

[0039] By adopting the above technical solution, by collecting relevant data on infrastructure construction, determining key indicators and calculating progress values, the comprehensive progress index is obtained through standardization and exponential operation, and then inversely transformed into a comprehensive progress value. This can fully and accurately reflect the progress of infrastructure construction in the area to be studied. It is helpful to better adjust the model parameters, improve the accuracy of drought and flood prediction, and provide a basis for the linkage of smart home devices.

[0040] Optionally, if the change information is an ecological environment change value, obtaining the change information of the area to be studied includes:

[0041] Collect and obtain ecological environment-related data, including changes in vegetation coverage, water quality indicators, soil quality indicators, and air quality indicators;

[0042] For each ecological environment indicator, calculate the change value between different time points, multiply the change value of each indicator by its corresponding weight, and get the weighted change value. Multiply the change value of the indicator by the preset weight to get the weighted change value as the ecological environment change value.

[0043] By adopting the above technical solution, a variety of ecological environment-related data are collected, and the change value of each indicator is calculated and weighted to obtain the ecological environment change value. It can fully reflect the changes in the regional ecological environment and provide a basis for adjusting the model parameters. It is helpful to improve the accuracy of drought and flood prediction and make the linkage of smart home devices more reasonable.

[0044] The optional, preset drought and flood risk level analysis algorithm is as follows:

[0045] Analyze and extract comprehensive risk index related indicators from the forecast results of various meteorological elements, geographic information data and other data. The comprehensive risk index related indicators include precipitation anomaly percentage, soil moisture index, river water level index and vegetation index change rate;

[0046] According to the corresponding relationship between the indicators and weights of different comprehensive risk indices, the weight of each indicator is analyzed and determined;

[0047] According to the weight of each indicator and the related indicators of the comprehensive risk index, the comprehensive risk index calculation formula is used to calculate the comprehensive risk index, which is as follows: ;

[0048] Among them, the precipitation anomaly percentage weight is , the precipitation anomaly percentage is , the soil moisture index weight is , is the soil moisture index, and the river water level index weight is , is the river water level index, and the vegetation index change rate is , the vegetation index change rate weight is , is the comprehensive risk index;

[0049] The drought or flood risk level is analyzed and determined based on the correspondence between the range in which the comprehensive risk index falls and the drought or flood risk level.

[0050] By adopting the above technical solution, multiple relevant indicators are extracted to comprehensively consider meteorological, geographical and other factors. Determine the indicator weights and calculate the comprehensive risk index to make the assessment more scientific and accurate. Determine the drought and flood risk level based on the comprehensive risk index, which can provide a clear basis for drought and flood warning, help take countermeasures in advance and reduce disaster losses. At the same time, it provides a reference for the linkage solution of smart home devices, improves the ability of the region to cope with drought and flood disasters, and promotes the stable development and ecological protection of the region.

[0051] In the second aspect, the present application provides a high temporal and spatial resolution drought-flood conversion prediction and warning system based on WRF, which adopts the following technical solutions:

[0052] A high spatiotemporal resolution drought-flood conversion prediction and warning system based on WRF, comprising a memory, a processor, and a program stored in the memory and executable on the processor, wherein the program can be loaded and executed by the processor to implement the high spatiotemporal resolution drought-flood conversion prediction and warning method based on WRF as described in the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a flow chart of a high spatiotemporal resolution drought-flood conversion prediction and early warning method based on WRF in an embodiment of the present application.

[0054] Figure 2 It is a flowchart diagram of another embodiment of the present application that analyzes and determines the model parameter setting scheme for the area to be studied according to the correspondence between the area category and the model parameter setting scheme, and executes the model parameter setting scheme for the area to be studied.

[0055] Figure 3 It is a flowchart diagram of another embodiment of the present application that analyzes and determines the model parameter setting scheme for the area to be studied according to the correspondence between the area category and the model parameter setting scheme, and executes the model parameter setting scheme for the area to be studied. DETAILED DESCRIPTION

[0056] The present application is further described in detail below in conjunction with the accompanying drawings.

[0057] Reference Figure 1 , is a high temporal and spatial resolution drought and flood conversion prediction and early warning method based on WRF disclosed in this application, comprising:

[0058] Step S100, obtaining meteorological data, geographic information data and other data, and performing data preprocessing.

[0059] Meteorological data: refers to various parameter information describing the state and changes of the atmosphere, including temperature, air pressure, humidity, wind speed, wind direction, precipitation, etc. These data reflect the physical characteristics and dynamic changes of the atmosphere and are crucial for drought and flood prediction.

[0060] Meteorological data can be obtained in the following ways: 1. Meteorological observation stations: Real-time observations are carried out through ground-based meteorological observation stations, high-altitude detection stations and other equipment to obtain data on conventional meteorological elements such as temperature, air pressure, and humidity. 2. Satellite remote sensing: Meteorological satellites can be used to obtain a wide range of meteorological data, such as cloud coverage, precipitation distribution, etc. For example, satellite cloud images can be used to analyze the movement and development trends of clouds, thereby inferring the possibility of precipitation. 3. Numerical weather forecast products: Numerical weather forecast products are obtained from professional meteorological agencies. These products usually contain rich meteorological data and can be used as a reference for drought and flood prediction.

[0061] Geographic information data: covers topography, land use type, water system distribution, etc. Topography affects atmospheric circulation and precipitation distribution; land use type determines the evaporation and infiltration of water on the surface; water system distribution is closely related to the storage and flow of water resources.

[0062] The ways to obtain geographic information data are as follows: 1. Geographic Information System (GIS): GIS software can be used to obtain and process various geographic information data. For example, topographic data, land use data, etc. can be downloaded from the GIS database. 2. Satellite images: Satellite images can be used to obtain high-resolution geographic information, such as land cover types, water system distribution, etc. After image interpretation and classification processing, the required geographic information data can be extracted.

[0063] Other data: may include historical records of drought and flood disasters, socio-economic data, etc. Historical records of drought and flood disasters help to understand the occurrence patterns and characteristics of regional drought and flood; socio-economic data such as population density and industrial structure can reflect the impact of drought and flood on different regions.

[0064] Step S200, obtaining relevant parameter information of the area to be studied, and matching and analyzing the area category of the area to be studied according to the parameter information associated with different categories of areas.

[0065] Related parameter information: refers to various characteristic parameters that can be used to distinguish different regional categories. These parameters may include geographical characteristic parameters (such as topography, altitude, longitude and latitude, etc.), climate characteristic parameters (such as annual average temperature, annual precipitation, seasonal distribution of precipitation, etc.), land use characteristic parameters (such as cultivated land area ratio, forest coverage, construction land ratio, etc.) and socio-economic characteristic parameters (such as population density, industrial structure, etc.).

[0066] The geographic feature parameters are obtained as follows:

[0067] Geographic Information System (GIS): GIS software can be used to obtain topographic data, altitude data, and longitude and latitude information. By analyzing the digital elevation model (DEM), the topographic features can be determined.

[0068] Global Positioning System (GPS): In field surveying, GPS can accurately determine the longitude and latitude coordinates of a specific location.

[0069] The acquisition of climate characteristic parameters is as follows: 1. Meteorological observation stations: Through meteorological observation stations distributed in different regions, climate data such as temperature and precipitation can be monitored for a long time. These observation stations will regularly record and upload meteorological data for analysis. 2. Satellite remote sensing: Meteorological satellites can be used to obtain a wide range of climate data, such as cloud distribution, precipitation, etc. By analyzing and processing satellite data, more comprehensive climate characteristic information can be obtained.

[0070] The acquisition of land use characteristic parameters is as follows: Satellite image interpretation: Through the analysis and interpretation of high-resolution satellite images, different land use types can be distinguished, such as cultivated land, forest land, construction land and water area, etc. Then, the area ratio of each type of land is calculated.

[0071] The socio-economic characteristic parameters are obtained as follows: Statistical department data: Data such as population density and economic development level can be obtained from national or local statistical departments.

[0072] Regional categories: The study area is divided into different categories according to different characteristic parameters, so that different model parameter setting schemes and drought and flood risk assessment methods can be used for different categories of regions. Regional categories can be divided according to actual conditions, for example, they can be divided into mountainous areas, plain areas, coastal areas, urban areas, rural areas, etc.

[0073] Take a prefecture-level city in a certain province as an example. First, the topographic data of the region was obtained through GIS software, and it was found that the region was mainly plains and low mountains and hills. Using the data statistics of the meteorological observation station, the average annual temperature in the region is 18℃, the annual precipitation is 1200 mm, and the precipitation is mainly concentrated in summer. Through the interpretation of satellite images, it is determined that the proportion of cultivated land in the region is 40%, the proportion of forest land is 30%, the proportion of construction land is 20%, and the proportion of water area is 10%. The data obtained from the statistical department showed that the population density in the region is 500 people per square kilometer, and the industrial structure is mainly agriculture, manufacturing and service industries.

[0074] Based on these parameter information, the region can be preliminarily divided into a plain and hilly mixed agricultural and industrial area. In the subsequent drought and flood conversion forecast and early warning, the model parameter setting scheme and drought and flood risk assessment method suitable for the plain and hilly terrain, agriculture and industry characteristics can be adopted according to the category of the region.

[0075] Step S300: analyzing and determining the model parameter setting scheme for the area to be studied according to the correspondence between the area category and the model parameter setting scheme, and executing the model parameter setting scheme for the area to be studied.

[0076] Model parameter setting scheme: A specific combination of parameters set for the WRF (Weather Research and Forecasting) model for different regional categories. These parameters include physical process parameters (such as radiation transfer, microphysical processes, boundary layer schemes, etc.), grid resolution, simulation time step, etc.

[0077] Establish the correspondence between regional categories and model parameter setting schemes: By analyzing the historical meteorological data and drought and flood events in different regions, summarize the differences in drought and flood formation mechanisms and meteorological characteristics of different regional categories. Based on these differences, determine the appropriate WRF model parameter setting scheme for different regional categories. For example, mountainous areas may require finer terrain resolution and consideration of the impact of terrain on airflow; coastal areas need to consider the regulation of the ocean on climate.

[0078] Determine the model parameter setting scheme for the area to be studied: Once the regional category of the area to be studied is determined, the corresponding model parameter setting scheme can be selected based on the pre-established correspondence.

[0079] Execute the model parameter setting scheme: Apply the determined model parameter setting scheme to the WRF model, including setting physical process parameters, adjusting grid resolution and simulation time step, etc. Ensure that the model can accurately simulate the atmospheric movement and change process in the area to be studied, and provide a reliable basis for subsequent drought and flood prediction.

[0080] Step S400, using a data assimilation algorithm to fuse data from different sources together to generate initial field data, and according to the correspondence between the condition information of the area to be studied and the boundary condition type, matching and obtaining the boundary condition type of the area to be studied, and using a boundary layer parameterization scheme to process the boundary layer.

[0081] Among them, the data assimilation algorithm is a technology that combines observational data with numerical models, and improves the accuracy and reliability of the model by continuously adjusting the model state to minimize the difference between the observed value and the model prediction value. It can integrate data from different observation methods (such as weather stations, satellites, radars, etc.) to provide a more accurate initial field for the model.

[0082] Initial field data: It is the initial state data required by the numerical model at the beginning of the simulation, including the distribution of meteorological elements such as atmospheric temperature, humidity, air pressure, wind field, etc. Accurate initial field data is crucial to the simulation results of the model.

[0083] Boundary condition type: refers to the specific type used in numerical models to describe the changes in meteorological elements on the model boundary. Common boundary condition types include fixed boundary conditions (such as given fixed meteorological element values), periodic boundary conditions (suitable for simulating periodic areas), etc. Different boundary condition types are suitable for different conditions in the area to be studied.

[0084] Boundary layer parameterization scheme: A mathematical method used to describe the physical processes in the atmospheric boundary layer. The atmospheric boundary layer is a layer of atmosphere close to the Earth's surface where the meteorological elements vary in a complex manner, including turbulence, heat and water vapor exchange processes. The choice of boundary layer parameterization scheme affects the model's simulation accuracy of the meteorological processes in the boundary layer.

[0085] Generate initial field data using data assimilation algorithms: Collect meteorological data and geographic information data from different sources, including meteorological observation station data, satellite remote sensing data, radar data, etc. Select appropriate data assimilation algorithms, such as three-dimensional variational method (3D-Var), four-dimensional variational method (4D-Var), etc. These algorithms continuously adjust the model state through optimization algorithms to minimize the difference between the model prediction value and the observation value. Combine the collected observation data with the WRF model, and fuse data from different sources through data assimilation algorithms to generate accurate initial field data.

[0086] Match and obtain the boundary condition type of the area to be studied: Analyze the condition information of the area to be studied, including geographical location, topography, climate characteristics, etc. According to the correspondence between different condition information and boundary condition types, select the boundary condition type suitable for the area to be studied. For example, for inland areas far away from the ocean, fixed boundary conditions may be used; for coastal areas or areas with obvious seasonal changes, more complex boundary condition types may be used.

[0087] Use boundary layer parameterization scheme to process the boundary layer: Select a suitable boundary layer parameterization scheme according to the characteristics of the area to be studied and the simulation requirements. Different parameterization schemes are suitable for different meteorological conditions and topography. Apply the selected boundary layer parameterization scheme to the WRF model to simulate and process the physical processes in the atmospheric boundary layer, and improve the model's simulation accuracy of the changes in meteorological elements in the boundary layer.

[0088] Assume that the area to be studied is an inland mountainous area. First, collect data from different sources such as meteorological observation station data and satellite remote sensing data in the area. Then, select the three-dimensional variational method as the data assimilation algorithm, combine these data with the WRF model, and generate initial field data.

[0089] For this inland mountainous area, fixed boundary conditions are selected as the boundary condition type according to its geographical location and terrain characteristics. In terms of boundary layer processing, considering the complex terrain and strong atmospheric turbulence in the mountainous area, a boundary layer parameterization scheme suitable for the mountainous area is selected, such as a turbulence parameterization scheme that considers the influence of terrain.

[0090] Step S500, the WRF model gradually advances the calculation according to the model parameter setting scheme and the initial conditions according to the time step to simulate the movement and change process of the atmosphere, wherein the initial conditions include the initial field setting and the boundary condition setting.

[0091] WRF model: Weather Research and Forecasting Model, is an advanced mesoscale numerical weather forecast model that can simulate various physical processes in the atmosphere, including atmospheric movement, water vapor transport, cloud physical processes, etc. It is used in weather forecasting, climate research, and atmospheric environment simulation.

[0092] Time step: The time interval between each calculation of the model during the simulation process. The selection of time step needs to consider the stability and computational efficiency of the model, and is usually adjusted according to the physical process and computing resources of the model.

[0093] Specifically, first load the initial conditions: load the initial field data generated in step S400 and the determined boundary condition type into the WRF model to provide the model with initial atmospheric state and boundary constraints. Ensure the accuracy of the initial field data and the rationality of the boundary conditions to ensure the reliability of the model simulation.

[0094] Setting model parameters: According to the model parameter setting scheme determined in step S300, various parameters of the WRF model are set, including physical process parameters, grid resolution, time step, etc. Different parameter settings will affect the simulation accuracy and calculation efficiency of the model, and need to be reasonably adjusted according to the characteristics and needs of the area to be studied.

[0095] Step-by-step calculation: According to the set time step, the WRF model gradually advances the calculation to simulate the movement and change of the atmosphere. In each time step, the model calculates the changes in the temperature, humidity, air pressure, wind field and other meteorological elements of the atmosphere at the next moment based on the current atmospheric state and physical process equations. By repeating this process continuously, the model can simulate the dynamic changes of the atmosphere over a period of time.

[0096] Considering physical processes: During the simulation process, the WRF model considers various atmospheric physical processes, such as radiation transfer, water vapor phase change, atmospheric turbulence, etc. Accurate simulation of these physical processes is crucial for predicting weather and climate phenomena, and different physical process parameterization schemes will affect the model's simulation accuracy of these processes.

[0097] Step S600: After the WRF model runs, it outputs the prediction results of various meteorological elements, and evaluates and determines the drought and flood risk level based on the output prediction results of various meteorological elements and the preset drought and flood risk level analysis algorithm.

[0098] Forecast results of meteorological elements: The WRF model outputs various parameters about the state of the atmosphere after the model runs, including temperature, air pressure, humidity, wind speed, wind direction, precipitation, etc. These forecast results reflect the possible state of the atmosphere in the future. Obtain forecast results of meteorological elements: After the WRF model runs, collect the forecast data of various meteorological elements. These data include information such as temperature, precipitation, humidity, etc. at different time points and spatial locations. These forecast results are sorted and analyzed to extract key meteorological elements related to drought and flood, such as precipitation, soil moisture, etc.

[0099] Drought and flood risk level analysis algorithm: A method used to assess the possibility and severity of drought and flood disasters in a specific area. It is usually based on a comprehensive analysis of multiple meteorological elements and geographic information data to determine the level of drought and flood risk.

[0100] Drought and flood risk level: A level that classifies the possibility and severity of drought and flood disasters, such as low risk, medium risk, high risk, etc. Different risk levels correspond to different response measures and warning levels.

[0101] Suppose that after the WRF model runs, a region obtains the forecast results of meteorological elements for the next week. Among them, the predicted precipitation value is much higher than the historical average precipitation for the same period, the soil moisture is also high, the river water level is at a normal level, and the vegetation index does not change much. According to the drought and flood risk level analysis algorithm, first calculate the relevant indicators such as the precipitation anomaly percentage and soil moisture index. Assuming that the precipitation anomaly percentage is high, the soil moisture index is also at a high level, while the river water level index and vegetation index change rate are relatively low. According to the preset weights, calculate the comprehensive risk index.

[0102] Step S700: confirm the smart home device linkage plan according to the correspondence between the drought and flood risk level and the smart home device linkage plan, and execute the smart home device linkage plan.

[0103] Smart home device linkage solution: A series of collaborative operation strategies for smart home devices are formulated according to different drought and flood risk levels, aiming to improve the family's ability to cope with drought and flood disasters and the convenience and safety of life.

[0104] Establish a correspondence: Predetermine the corresponding rules between drought and flood risk levels and smart home device linkage solutions. For example, a low risk level may correspond to the normal operation of equipment, such as the normal operation of smart ventilation equipment to maintain indoor air circulation; a medium risk level may trigger some early warning equipment, such as smart flood sensors to start enhanced monitoring; a high risk level may initiate more comprehensive protective measures, such as automatically closing smart windows and starting smart drainage pumps.

[0105] Confirm linkage scheme: After the drought and flood risk level of the area to be studied is determined, the linkage scheme of smart home devices suitable for the risk level is found and confirmed according to the pre-established correspondence. For example, if the drought and flood risk level is assessed as medium risk, then the corresponding medium risk linkage scheme is determined, which may include starting the high-sensitivity monitoring mode of the smart water sensor, notifying the user of the smart voice assistant to pay attention to weather changes, etc.

[0106] Execute linkage plan: Once the linkage plan of smart home devices is confirmed, the smart home control system sends instructions to each related device to execute the linkage plan. After receiving the instructions, the smart home devices act according to the predetermined operation. For example, the smart water sensor starts to monitor the ground humidity at a high frequency, and the smart voice assistant sends weather warnings and response suggestions to the user.

[0107] Assume that a family has installed a smart home system, and the system is connected to the drought and flood risk warning mechanism. When the WRF model predicts that the drought and flood risk level in the area is low, the smart home devices operate according to the normal settings, the smart air conditioner maintains a comfortable temperature and humidity in the room, and the smart lights automatically adjust the brightness according to the indoor light.

[0108] If the drought or flood risk level rises to medium risk, the smart home system will automatically activate the smart flood sensor to strengthen monitoring of areas where flooding may occur. At the same time, the smart voice assistant will send reminders to users to inform them of the current drought or flood risk situation and suggest that users check the drainage facilities of their homes.

[0109] When the drought or flood risk level reaches a high risk, the smart home system will automatically close the smart windows to prevent rain from entering the room. At the same time, the smart drainage pump will be started to remove any accumulated water in time. In addition, the smart energy management system will adjust the operating status of household electrical equipment to ensure a stable power supply during disasters.

[0110] Reference Figure 2 According to the corresponding relationship between the regional category and the model parameter setting scheme, the model parameter setting scheme of the area to be studied is analyzed and determined, and the model parameter setting scheme of the area to be studied is implemented, including:

[0111] Step S310, analyzing whether there are extreme meteorological events in the area to be studied within a preset time period. If yes, execute step S320; if no, execute step S340.

[0112] Preset time period: A pre-set time range used to determine whether extreme meteorological events have occurred in the area to be studied. This time period can be determined based on specific research needs and the availability of meteorological data, such as one week, one month, one quarter, etc.

[0113] Extreme weather events: refers to weather phenomena that are beyond the normal range of weather changes and have a large impact and potential harm, such as heavy rain, heavy snow, high temperature heat waves, low temperature cold waves, typhoons, tornadoes, etc. These events usually have a serious impact on people's lives, production and the natural environment.

[0114] The specific analysis process is as follows:

[0115] Data collection: Collect meteorological data of the area to be studied within a preset time period, including observation data of various meteorological elements such as temperature, precipitation, wind speed, and air pressure. These data can be obtained from meteorological observation stations, satellite remote sensing, numerical weather forecast products and other channels. At the same time, other information related to extreme meteorological events can also be collected, such as disaster reports, news reports, and information on social media, to assist in determining whether an extreme meteorological event has occurred.

[0116] Event judgment: Based on the collected meteorological data and related information, analyze whether the situation that meets the definition of extreme meteorological events has occurred in the study area within the preset time period. Some thresholds or standards can be set to determine whether an extreme meteorological event has occurred. For example, when the precipitation exceeds a certain multiple (such as twice) of the historical average for the same period, it can be considered that a heavy rainstorm extreme meteorological event has occurred; when the temperature is higher or lower than the historical average for a certain range for many consecutive days, it can be considered that a high temperature heat wave or a low temperature cold wave extreme meteorological event has occurred.

[0117] Step S320, obtaining data on historical extreme meteorological events, classifying extreme meteorological events according to the data on historical extreme meteorological events, and extracting characteristic parameters of the events.

[0118] Data on historical extreme weather events: Information on extreme weather events that occurred in the past, including the time and location of the event, weather element data (such as precipitation, wind speed, temperature, etc.), the scope and extent of the disaster impact, etc. These data can be obtained from meteorological archives, disaster databases, news reports, and other channels.

[0119] Characteristic parameters: Specific parameters used to describe the characteristics of extreme meteorological events. These parameters can help identify and distinguish different extreme events, and also provide a basis for subsequent model parameter settings. Characteristic parameters can include the intensity, duration, impact range, peak value of meteorological elements, etc.

[0120] Obtain historical data: Collect historical extreme weather event data in the area to be studied and surrounding areas. This data can be obtained from meteorological departments, scientific research institutions, databases and other channels. Ensure the accuracy and completeness of the data, including the time, location, and meteorological element data of the event. Sort and filter the collected data to remove irrelevant or inaccurate data.

[0121] Event classification: Classify historical extreme meteorological events according to predetermined classification criteria. For example, rainstorm and flood events can be classified into one category, and high temperature and heat wave events can be classified into another category based on the meteorological characteristics of the events. The classification process can adopt a combination of manual classification and automatic classification. Manual classification can be judged based on experience and professional knowledge, and automatic classification can use data mining and machine learning algorithms, such as cluster analysis, to quickly classify large amounts of historical data.

[0122] Extract characteristic parameters: For each type of extreme meteorological event, analyze its characteristics and extract corresponding characteristic parameters. For example, for rainstorm and flood events, parameters such as peak precipitation, duration, and impact range can be extracted; for high temperature heat wave events, parameters such as maximum temperature, duration, and impact area can be extracted. The extraction of characteristic parameters can be achieved through statistical analysis and calculation of historical data. For example, calculate the average precipitation and maximum temperature during the event to determine the duration and impact range of the event.

[0123] Step S330, input the feature parameters of the extracted event into the trained model parameter setting matching model, and output the model parameter setting scheme.

[0124] Model parameter setting matching model: A trained algorithm model that can output a model parameter setting scheme suitable for a specific extreme weather event based on the input characteristic parameters. It is usually obtained by learning and training a large number of historical extreme weather events and their corresponding model parameter settings using machine learning and other technologies.

[0125] Specifically, the process is as follows:

[0126] Prepare input data: Arrange and format the characteristic parameters of the extreme weather event extracted in step S320 to meet the input requirements of the model parameter setting matching model. For example, convert the characteristic parameters into a specific data format, or perform normalization processing, etc.

[0127] Input feature parameters: Input the sorted feature parameters into the trained model parameter setting matching model. This model can be built based on machine learning algorithms such as neural networks, decision trees, support vector machines, etc.

[0128] Model calculation and output: The model performs calculations and inferences based on the input feature parameters, and outputs a model parameter setting plan suitable for the current extreme meteorological event by matching and analyzing the learned historical data patterns.

[0129] The output model parameter setting scheme may include specific values ​​of various parameters, such as the grid resolution is set to a certain number of kilometers, the physical process parameter scheme is adopted, the simulation time step is set to a certain number of minutes, etc.

[0130] Step S340, maintaining the original settings.

[0131] Reference Figure 3 According to the corresponding relationship between the regional category and the model parameter setting scheme, the model parameter setting scheme of the area to be studied is analyzed and determined, and the model parameter setting scheme of the area to be studied is implemented, including:

[0132] Step S3A0, obtaining change information of the area to be studied, the change information including comprehensive land use change value, comprehensive infrastructure construction progress value, and ecological environment change value.

[0133] Comprehensive value of land use change: a comprehensive indicator reflecting the change of land use in the area to be studied. It is obtained by analyzing and calculating the changes in the area of ​​land use types in different periods, including the changes in different land use types such as cultivated land, forest land, construction land, and water areas.

[0134] Comprehensive infrastructure construction progress value: a value that measures the progress of infrastructure construction in the area to be studied. It covers infrastructure construction in many aspects, such as the length of road construction, building area, and number of public facilities.

[0135] Ecological environment change value: an indicator that indicates the degree of ecological environment change in the area to be studied. It takes into account changes in multiple ecological environment factors such as vegetation coverage, water quality indicators, soil quality indicators, and air quality indicators.

[0136] Step S3B0, analyzing whether each type of value of the change information exceeds the preset change threshold of the corresponding category. If yes, execute step S3C0; if no, execute step S3D0.

[0137] Various values ​​of change information: refer to the specific numerical values ​​corresponding to three different types of change information: comprehensive value of land use change value, comprehensive infrastructure construction progress value, and ecological environment change value.

[0138] Determine the preset change thresholds for various types of change information:

[0139] For the comprehensive value of land use change, a value indicating a significant land use change can be determined as a preset change threshold based on historical data and experience. For example, if the comprehensive value of land use change exceeds a certain proportion, such as 20%, it is considered that a significant change has occurred in land use.

[0140] For the comprehensive infrastructure construction progress value, a value indicating a faster construction progress can be determined as a preset change threshold based on the scale and speed of infrastructure construction. For example, when the comprehensive progress values ​​of road construction length, building floor area, number of public facilities construction, etc. increase by more than a certain amount within a certain period of time, it is considered to exceed the preset change threshold.

[0141] For the ecological environment change value, a value indicating a large ecological environment change can be determined as a preset change threshold in combination with the ecological environment quality standard and sensitivity analysis. For example, when the comprehensive change value of vegetation coverage change, water quality index, soil quality index, air quality index, etc. exceeds a certain range, it is judged to exceed the preset change threshold.

[0142] Step S3C0, taking the regional variation characteristics as input objects, inputting them into the relationship model between the trained physical process parameterization scheme and the regional variation characteristics, and outputting the adjustment parameters of the physical process parameterization scheme.

[0143] Regional change characteristics: It is composed of the comprehensive value of land use change, the comprehensive infrastructure construction progress value, the ecological environment change value, etc., reflecting the changes in land use, infrastructure construction and ecological environment in the study area.

[0144] Relationship model: A model that can establish the relationship between the physical process parameterization scheme and the regional change characteristics after training with a large amount of data. It can predict the adjustment parameters of the physical process parameterization scheme suitable for the region based on the regional change characteristics.

[0145] Adjustment parameters of the physical process parameterization scheme: used to adjust the specific values ​​of the physical process parameterization scheme in the WRF model to make the model better adapt to the atmospheric movement and change process after regional changes, and improve the accuracy of drought and flood prediction.

[0146] Model calculation and output: The relational model performs calculations and inferences based on the input regional change characteristics. By analyzing the impact of regional changes on atmospheric physical processes and combining similar situations in historical data, the adjustment parameters of the physical process parameterization scheme suitable for the current regional changes are determined. The adjustment parameters are output, which may include adjustment values ​​of radiation transfer parameters, microphysical process parameters, boundary layer scheme parameters, etc. For example, if the land use change in the region causes the surface albedo to change, the relational model may output the value of the adjusted radiation transfer parameter to better simulate the solar radiation absorption and reflection after the regional change.

[0147] Step S3D0, analyzing and determining the model parameter setting scheme for the area to be studied according to the correspondence between the area category and the model parameter setting scheme, and executing the model parameter setting scheme for the area to be studied.

[0148] If the change information is the land use change value, the change information of the area to be studied includes:

[0149] Step S3A1, using satellite remote sensing images to obtain land use data of different expected study areas, and using image classification algorithms to divide the land into different use types, including cultivated land, forest land, construction land, and water area.

[0150] Satellite remote sensing images: image data of the earth's surface obtained by sensors carried by satellites. Different satellite sensors can obtain electromagnetic wave information in different bands to reflect various characteristics of the earth's surface.

[0151] Land use data: information describing how different areas of the earth’s surface are used for different purposes, such as arable land for agricultural planting, forest land for forestry production, and construction land for urban construction.

[0152] Image classification algorithm: A method that uses computer technology to analyze and process satellite remote sensing images and classify each pixel in the image into different land use types. This application selects a suitable image classification algorithm, such as supervised classification (such as maximum likelihood method, support vector machine, etc.) or unsupervised classification (such as ISODATA algorithm, K-Means algorithm, etc.). If supervised classification is used, it is necessary to first select some representative training samples. The training samples are pixels or areas of known land use types, which are used to train the classifier. By analyzing the features of the training samples, the classifier learns the characteristic patterns of different land use types. For unsupervised classification, the algorithm automatically divides the pixels in the image into different categories, and then names and explains these categories based on prior knowledge or field surveys. Run the image classification algorithm to divide each pixel in the satellite remote sensing image into different land use types such as cultivated land, forest land, construction land, and water areas.

[0153] Step S3A2, compare and analyze the land use classification results of different periods, and calculate the change area ratio and change speed of each land use type.

[0154] Specifically, we first sorted out the land use classification results of different periods to ensure the accuracy and completeness of the data, and aligned the classification results of different periods according to the same coordinate system and spatial resolution for comparative analysis.

[0155] For each land use type, count the area in different periods. For example, for cultivated land, count the cultivated land area at two time points. Calculate the changed area of ​​each land use type between the two time points. The changed area is equal to the area of ​​the later period minus the area of ​​the previous period. If the changed area is a positive value, it means that the area of ​​this type has increased; if it is a negative value, it means that the area has decreased.

[0156] Calculate the proportion of area that changed as follows: For each land use type, calculate the proportion of area that changed. The proportion of area that changed is equal to the area that changed divided by the total area of ​​that type in the previous period. For example, if the total area of ​​cultivated land in the previous period was 100 square kilometers, and the area that changed between the two time points was -10 square kilometers (a decrease in area), then the proportion of area that changed in cultivated land is -10%.

[0157] Calculate the rate of change: Determine the time interval, usually in years. If the interval between two time points is 5 years, then when calculating the rate of change, divide the area of ​​change by the time interval (5 years) to get the average area of ​​change per year. For each land use type, the rate of change is equal to the average area of ​​change per year divided by the total area of ​​that type in the previous period. For example, if the average annual change in cultivated land area is -2 square kilometers / year, and the total area in the previous period is 100 square kilometers, then the rate of change of cultivated land is -2% / year.

[0158] Step S3A3, standardize the two indicators of the change area ratio and the change speed, and perform weighted summation on the standardized change area ratio and change speed according to the importance weights of different indicators to obtain the comprehensive change value of each land use type.

[0159] Standardization: The values ​​of the two indicators, the proportion of change area and the speed of change, are converted into numerical values ​​within a specific range for comparison and weighted summation. Standardization can eliminate the impact of different indicator dimensions and make them comparable.

[0160] Standardization: For the change area ratio, assume that its value range is [-a, b], where a and b are positive numbers. The linear standardization method can be used to map the value of the change area ratio to the interval [0, 1]. The specific formula is: Standardized change area ratio = (original change area ratio + a) / (a ​​+ b). For the change speed, a similar method can be used for standardization. Assume that the change speed ranges from [-c, d], where c and d are positive numbers. Standardized change speed = (original change speed + c) / (c + d).

[0161] Determine the importance weights: Determine the importance weights of the change area ratio and the change speed based on expert experience, historical data analysis, or other methods. The importance weights range from [0,1], and the sum of the two weights is 1. For example, the importance weight of the change area ratio can be considered as w1, the importance weight of the change speed can be considered as w2, and w1+w2=1.

[0162] Weighted summation: For each land use type, multiply the standardized change area ratio by its importance weight w1, and multiply the standardized change rate by its importance weight w2. Add the two products to get the comprehensive change value of each land use type. The calculation formula for the comprehensive change value is: Comprehensive change value = standardized change area ratio × w1 + standardized change rate × w2.

[0163] Step S3A4, summing up the comprehensive change values ​​of all land use types to obtain a comprehensive land use change value.

[0164] Comprehensive change value: The value obtained by standardizing the change area ratio and change speed of each land use type and weighted summing them up according to the importance weight reflects the comprehensive change of the land use type.

[0165] Comprehensive value of land use change: The value obtained by summing up the comprehensive change values ​​of all land use types is used to measure the degree of land use change in the study area as a whole.

[0166] If the change information is the progress value of comprehensive infrastructure construction, the change information of the area to be studied includes:

[0167] Step S3AA, collect and obtain infrastructure construction related data of the area to be studied, the infrastructure construction related data includes road construction length, building construction area, and the number of public facilities construction.

[0168] The road construction length data can use geographic information system (GIS) data: if the city management department or relevant institutions in the area have a public GIS data platform that contains road information layers, it can be connected and data extracted through the data interface of these platforms. Use relevant GIS software or development tools to write programs to obtain the geometric information of the road, such as the line length of the road, the coordinates of the starting and ending points of the road, etc., and then calculate the road construction length based on this information.

[0169] The floor area of ​​a building can be obtained using the API of the Building Information Modeling (BIM) software. For construction projects designed and managed using BIM technology, the BIM software contains detailed information about the building, including floor area, floor layout, building components, etc. You can use the application programming interface (API) provided by the BIM software to write a program to access and extract the floor area data of the building.

[0170] The number of public facility construction can be obtained through pre-configured intelligent detection equipment for monitoring public facilities.

[0171] Step S3AB, based on the correspondence between different infrastructure construction categories and key indicators, determine the key indicators and calculate the progress value of the infrastructure construction.

[0172] Key indicators: Specific indicators used to measure the progress of different infrastructure construction categories, such as the increase in the length of road construction, the growth in the floor area of ​​buildings, the increase in the number of public facilities construction, etc.

[0173] For the road construction category, the key indicator can be the increase in the length of road construction, for example, the length of newly built roads within a certain period of time or the length of widening and reconstruction of existing roads.

[0174] For the building construction category, the key indicator can be the growth of building floor area, including the floor area of ​​new buildings and the floor area increased by expansion and renovation of existing buildings.

[0175] For the public facilities construction category, the key indicator can be the increase in the number of public facilities construction, such as the number of newly added schools, hospitals, parks, stadiums and other public facilities.

[0176] Calculate key indicator values:

[0177] Calculation of the increase in road construction length: Obtain the road construction length data at different time points, subtract the road construction length at the previous time point from the road construction length at the later time point, and get the increase in road construction length. For example, if the road construction length in the study area is 100 kilometers at the beginning of the year and 120 kilometers at the end of the year, then the increase in road construction length is 20 kilometers.

[0178] Calculation of the growth of building floor area: Similarly, obtain the building floor area data at different time points and subtract them to get the growth. For example, the building floor area of ​​the study area was 500,000 square meters last month and 550,000 square meters this month, and the growth of building floor area is 50,000 square meters.

[0179] Calculation of the increase in the number of public facilities: Count the number of various public facilities at different time points and calculate the difference. For example, the number of schools in the study area was 10 last quarter, and it increased to 12 this quarter, and the increase in the number of public facilities is 2.

[0180] Calculate the progress value of infrastructure construction: assign weights to each key indicator based on the importance of different infrastructure construction categories to the overall infrastructure construction. For example, the weight of road construction is w1, the weight of building construction is w2, and the weight of public facility construction is w3, and w1+w2+w3=1. The progress value of infrastructure construction = the increase in the length of road construction × w1 + the increase in the construction area of ​​buildings × w2 + the increase in the number of public facility construction × w3.

[0181] Step S3AC, standardize the progress value of each infrastructure construction item, and perform exponential calculation on the standardized single progress value to obtain the single progress index.

[0182] Single progress index: The value obtained by performing an index operation on the standardized infrastructure construction progress value, which is used to measure the relative progress of a specific infrastructure construction category.

[0183] Standardization: Convert the infrastructure construction progress values ​​of different dimensions into a unified and comparable numerical range for subsequent index calculation and comprehensive calculation. The specific standardization process is as follows: First, determine the standardization range. Suppose we standardize the infrastructure construction progress value to the interval [0,1]. For the road construction progress value, assume that its value range is [a1,b1]. The standardization formula is: Standardized road construction progress value = (original road construction progress value-a1) / (b1-a1). For example, if the road construction progress value range is [0,50] kilometers, and the road construction progress value at a certain moment is 20 kilometers, then the standardized road construction progress value = (20-0) / (50-0)=0.4. For the building construction area progress value, assume that its value range is [a2,b2]. The standardization formula is similar: Standardized building construction area progress value = (original building construction area progress value-a2) / (b2-a2). For the public facility construction quantity progress value, assume that its value range is [a3,b3]. The standardization formula is: The standardized public facilities construction quantity progress value = (original public facilities construction quantity progress value - a3) / (b3-a3).

[0184] The index operation is as follows: perform index operation on the standardized progress values ​​of various infrastructure construction projects. For example, for the standardized progress value of road construction r, the single progress index (road) = exp(r), where exp represents an exponential function with the natural constant e as the base. Similarly, perform index operation on the standardized progress values ​​of building construction area and the progress values ​​of the number of public facilities construction to obtain the single progress index (building) and the single progress index (public facilities).

[0185] Assume that the progress value range of road construction is [0,40] kilometers, and the progress value at a certain moment is 15 kilometers. The progress value range of building construction area is [0,10000] square meters, and the progress value at a certain moment is 3000 square meters. The progress value range of the number of public facilities construction is [0,10] units, and the progress value at a certain moment is 4 units.

[0186] Standardization: The standardized road construction progress value = (15-0) / (40-0) = 0.375. The standardized building construction area progress value = (3000-0) / (10000-0) = 0.3. The standardized public facility construction quantity progress value = (4-0) / (10-0) = 0.4.

[0187] Exponential calculation: Single progress index (road) = exp(0.375). Single progress index (building) = exp(0.3). Single progress index (public facilities) = exp(0.4).

[0188] Step S3AD, weighted sum of each individual progress index to obtain a comprehensive infrastructure construction progress index, and then obtain a comprehensive infrastructure construction progress value through index inverse transformation.

[0189] Comprehensive infrastructure construction progress index: a value that comprehensively reflects the progress of the overall infrastructure construction in the studied area, obtained by weighted summation of the individual progress indexes. Index inverse transformation: the process of converting the comprehensive infrastructure construction progress index into the actual comprehensive infrastructure construction progress value.

[0190] The specific implementation process of step S3AD is as follows: 1. Determine the weight: According to the actual situation and needs of the area to be studied, determine the importance weights of different types of infrastructure construction such as road construction, building construction, and public facilities construction in the comprehensive evaluation. For example, assume that the weight of road construction is w1, the weight of building construction is w2, and the weight of public facilities construction is w3, and w1+w2+w3=1. 2. Weighted sum to obtain the comprehensive infrastructure construction progress index: multiply each individual progress index by its corresponding weight, and then perform the sum operation. Assuming that the single progress index (road) is I1, the single progress index (building) is I2, and the single progress index (public facilities) is I3, then the comprehensive infrastructure construction progress index = I1×w1+I2×w2+I3×w3. 3. Index inverse transformation to obtain the comprehensive infrastructure construction progress value: perform index inverse transformation on the comprehensive infrastructure construction progress index to obtain the actual comprehensive infrastructure construction progress value. Assume that the comprehensive infrastructure construction progress index is I_total, and the exponential inverse transformation formula is comprehensive infrastructure construction progress value = ln(I_total), where ln represents the logarithmic function with the natural constant e as the base.

[0191] Assume that the single progress index (road) is 2.5, the single progress index (building) is 3, and the single progress index (public facilities) is 2.2. The road construction weight w1=0.4, the building construction weight w2=0.4, and the public facility construction weight w3=0.2.

[0192] The weighted summation gives the comprehensive infrastructure construction progress index: Comprehensive infrastructure construction progress index = 2.5 × 0.4 + 3 × 0.4 + 2.2 × 0.2 = 1 + 1.2 + 0.44 = 2.64. The index inverse transformation gives the comprehensive infrastructure construction progress value: Comprehensive infrastructure construction progress value = ln (2.64) ≈ 0.97.

[0193] If the change information is the ecological environment change value, the change information of the area to be studied includes:

[0194] Step S3Aa, collect and obtain ecological environment related data, wherein the ecological environment related data include vegetation coverage changes, water quality indicators, soil quality indicators, and air quality indicators.

[0195] Step S3Ab, for each ecological environment indicator, calculate the change value between different time points, multiply the change value of each indicator by its corresponding weight, and obtain the weighted change value. The change value of the indicator is multiplied by the preset weight to obtain the weighted change value as the ecological environment change value.

[0196] The calculation index change value is as follows:

[0197] For the vegetation coverage change index, assuming that the vegetation coverage at time point T1 is V1 and the vegetation coverage at time point T2 is V2, the vegetation coverage change value = V2-V1. For example, if the vegetation coverage at T1 is 40% and at T2 is 35%, the vegetation coverage change value is -5%.

[0198] For water quality indicators, multiple key parameters can be selected for comprehensive consideration. For example, let the chemical oxygen demand (COD) of the water body at time point T1 be C1, the dissolved oxygen content be D1, and the COD at time point T2 be C2, and the dissolved oxygen content be D2. First, calculate the change values ​​of COD and dissolved oxygen respectively, COD change value = C2-C1, dissolved oxygen change value = D2-D1. Then, according to specific needs, a comprehensive calculation method (such as weighted average, etc.) can be used to obtain the change value of the water quality indicator. For example, if the COD change value is +10mg / L and the dissolved oxygen change value is -2mg / L, after comprehensive calculation, the change value of the water quality indicator is a certain value.

[0199] For soil quality indicators, multiple parameters can also be selected for analysis. For example, let the soil organic matter content at time point T1 be O1 and the pH value be P1, and let the soil organic matter content at time point T2 be O2 and the pH value be P2. Calculate the change value of organic matter content and pH value respectively, organic matter content change value = O2-O1, pH value change value = P2-P1. Then use appropriate methods to comprehensively obtain the change value of soil quality indicators.

[0200] For air quality indicators, the changes in the concentration of major pollutants can be considered. For example, if the PM 2.5 The concentration is M1, the sulfur dioxide (SO2) concentration is S1, and the PM at time point T2 2.5 The concentration is M2, and the SO2 concentration is S2. Calculate PM 2.5 Change value = M2-M1, SO2 change value = S2-S1, and then the change value of air quality index is obtained comprehensively.

[0201] Determine the weight of indicators as follows: The weight of each ecological environment indicator can be determined through expert evaluation, hierarchical analysis method and other methods. For example, invite experts in the field of ecological environment to score the importance of vegetation coverage changes, water quality indicators, soil quality indicators, and air quality indicators based on their experience and expertise, and then determine the weight of each indicator through statistical analysis. Suppose that after evaluation, the weight of vegetation coverage changes is determined to be w1, the weight of water quality indicators is w2, the weight of soil quality indicators is w3, and the weight of air quality indicators is w4, and w1+w2+w3+w4=1.

[0202] Calculate the weighted change value: multiply the change value of each ecological environment indicator by its corresponding weight to obtain the weighted change value. For example, if the change value of vegetation coverage is V_change and its weight is w1, then the weighted change value of vegetation coverage is V_change×w1. Similarly, calculate the weighted change values ​​of water quality indicators, soil quality indicators, and air quality indicators.

[0203] Determine the ecological environment change value: sum the weighted change values ​​of each ecological environment index to obtain the ecological environment change value. That is, the ecological environment change value = the weighted change value of vegetation coverage + the weighted change value of water quality index + the weighted change value of soil quality index + the weighted change value of air quality index.

[0204] Assume that after calculation, the vegetation coverage change value is -8%, the water quality index change value is -0.5 (the value after comprehensive calculation), the soil quality index change value is +0.3, and the air quality index change value is -0.2.

[0205] Assume that the weights are: vegetation coverage change weight w1=0.3, water quality index weight w2=0.4, soil quality index weight w3=0.2, air quality index weight w4=0.1. Then the weighted change values ​​are: vegetation coverage weighted change value = -8% × 0.3 = -2.4%. Water quality index weighted change value = -0.5 × 0.4 = -0.2. Soil quality index weighted change value = 0.3 × 0.2 = 0.06. Air quality index weighted change value = -0.2 × 0.1 = -0.02. Ecological environment change value = -2.4%-0.2+0.06-0.02=-2.56%.

[0206] The preset drought and flood risk level analysis algorithm is as follows:

[0207] Step SA00, analyze and extract comprehensive risk index related indicators from the forecast results of various meteorological elements, geographic information data and other data. The comprehensive risk index related indicators include precipitation anomaly percentage, soil moisture index, river water level index and vegetation index change rate.

[0208] Analyze the forecast results of meteorological elements: Extract precipitation-related information from the forecast results of various meteorological elements output by the WRF model. For example, count the predicted precipitation within a certain period of time, compare it with the historical average precipitation during the same period, and calculate the precipitation anomaly percentage. Precipitation anomaly percentage = (actual precipitation - average precipitation during the same period) / average precipitation during the same period × 100%. If the forecast results show that the precipitation in a certain area in the future is significantly higher or lower than the historical average level during the same period, then the precipitation anomaly percentage in the area will change significantly, which may indicate a change in the risk of drought and flood. At the same time, pay attention to the impact of other meteorological elements such as temperature on soil moisture and vegetation growth. Higher temperatures may lead to faster evaporation of soil moisture, affecting the soil moisture index and the rate of change of the vegetation index.

[0209] Utilize geographic information data: Combined with the topographic information in the geographic information data, analyze the precipitation distribution and runoff in different terrain areas to determine the changing trend of the river water level index. For example, in mountainous areas, precipitation easily forms runoff, which may cause the river water level to rise rapidly; while in plain areas, runoff is relatively slow and the river water level may change more steadily. Evaluate the soil moisture conditions in different regions based on information such as altitude and river distribution. Generally speaking, the soil moisture in areas with lower altitudes and close to rivers is relatively high, while the soil moisture in areas with higher altitudes and far from rivers may be lower.

[0210] Extract relevant information from other data: If there is soil moisture monitoring data, it can be used directly to calculate the soil moisture index. The soil moisture index can be determined by measuring the moisture content in the soil or obtaining soil moisture information using remote sensing technology. For example, the reflectivity of a specific band in satellite remote sensing data is used to estimate soil moisture. The lower the reflectivity, the higher the soil moisture. For the vegetation index change rate, the vegetation index data in satellite remote sensing images can be used for calculation. Common vegetation indices such as the Normalized Difference Vegetation Index (NDVI) calculate the vegetation index change rate by comparing the NDVI values ​​at different time points. If the vegetation index change rate is positive, it means that the vegetation is growing well and may have a certain buffering effect on drought and floods; if it is negative, it may mean that the vegetation is affected by disasters such as drought or floods.

[0211] Comprehensive analysis and extraction of indicators: Comprehensive analysis will be conducted on the information extracted from the meteorological element forecast results, geographic information data and other data to determine the comprehensive risk index related indicators such as precipitation anomaly percentage, soil moisture index, river water level index and vegetation index change rate. For example, if the predicted precipitation is high, and the geographic information shows that the area has a low terrain and many rivers, and the soil moisture monitoring data shows that the soil moisture is high, then it can be preliminarily judged that the precipitation anomaly percentage, river water level index and soil moisture index in the area may be high, and the vegetation index change rate may be affected to a certain extent.

[0212] Assuming that the area to be studied is a river basin, the meteorological element forecast results of the WRF model show that the precipitation in the river basin in the next week is expected to be 30% higher than the historical average precipitation in the same period, and the precipitation anomaly percentage is calculated to be 30%.

[0213] Using geographic information data, we analyzed the topography and river distribution of the basin. We found that most of the basin is plain with crisscrossing rivers. Based on the altitude and river distribution, we preliminarily determined that the soil moisture in areas close to rivers may be higher. Combined with soil moisture monitoring data, we determined that the soil moisture index in the area is 0.7 (assuming that the soil moisture index ranges from 0 to 1).

[0214] Through the analysis of river water level monitoring data, it was found that due to the recent heavy rainfall, the river water level is already 20% higher than the normal level. The calculated river water level index is 0.2 (assuming that the river water level index ranges from 0 to 1).

[0215] Using the NDVI data from satellite remote sensing images, we calculated that the vegetation index change rate in the basin over the past month was -5%, indicating that the vegetation growth condition has been affected to a certain extent.

[0216] Through the above analysis, the relevant indicators of the comprehensive risk index of the basin are extracted as follows: precipitation anomaly percentage 30%, soil moisture index 0.7, river water level index 0.2 and vegetation index change rate -5%. These indicators will be used to calculate the comprehensive risk index in the future to determine the drought and flood risk level.

[0217] Step SB00, analyze and determine the weight of each indicator based on the corresponding relationship between the indicators and weights related to different comprehensive risk indices.

[0218] Specifically, weight: indicates the importance of each indicator in the calculation of the comprehensive risk index. The larger the weight value, the greater the impact of the indicator on the comprehensive risk index. Expert evaluation method can be used.

[0219] Step SC00, according to the weight of each indicator and the related indicators of the comprehensive risk index, the comprehensive risk index calculation formula is used to calculate the comprehensive risk index calculation formula, which is as follows: .

[0220] Among them, the precipitation anomaly percentage weight is , the precipitation anomaly percentage is , the soil moisture index weight is , is the soil moisture index, and the river water level index weight is , is the river water level index, and the vegetation index change rate is , the vegetation index change rate weight is , It is a comprehensive risk index.

[0221] Step SD00, analyzing and determining the drought or flood risk level according to the corresponding relationship between the interval range that the comprehensive risk index falls into and the drought or flood risk level.

[0222] Define drought and flood risk levels: assign a drought and flood risk level to each interval. The risk level can be expressed in numbers or described in words. For example, the comprehensive risk index can be defined as a low risk level in the interval [0,0.2], a lower risk level in the interval (0.2,0.4], a medium risk level in the interval (0.4,0.6], a higher risk level in the interval (0.6,0.8], and a high risk level in the interval (0.8,1].

[0223] Determine the drought and flood risk level: Compare the calculated comprehensive risk index with the divided intervals to determine which interval the index falls into. According to the drought and flood risk level corresponding to the interval, determine the drought and flood risk level of the area to be studied. For example, if the comprehensive risk index is 0.45, it falls into the (0.4, 0.6] interval, and the corresponding drought and flood risk level is a medium risk level.

[0224] Based on the same inventive concept, an embodiment of the present invention provides a high temporal and spatial resolution drought and flood conversion prediction and warning system based on WRF, including a memory and a processor, wherein the memory stores a program that can be executed on the processor to implement the following Figures 1 to 3 Procedure for either method.

[0225] The embodiments of this specific implementation method are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Therefore, all equivalent changes made based on the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A high temporal and spatial resolution drought and flood conversion prediction and early warning method based on WRF, characterized in that: include: Obtain meteorological data, geographic information data and other data, and perform data preprocessing; Obtain relevant parameter information of the area to be studied, and match and analyze the area category of the area to be studied based on the parameter information associated with different categories of areas; According to the correspondence between the regional category and the model parameter setting scheme, the model parameter setting scheme of the area to be studied is analyzed and determined, and the model parameter setting scheme of the area to be studied is implemented; Use data assimilation algorithms to fuse data from different sources to generate initial field data, and match and obtain the boundary condition type of the area to be studied based on the correspondence between the condition information of the area to be studied and the boundary condition type, and use the boundary layer parameterization scheme to process the boundary layer; The WRF model simulates the movement and change of the atmosphere by gradually advancing the calculation according to the model parameter setting scheme and initial conditions according to the time step. The initial conditions include the initial field setting and the boundary condition setting. After the WRF model runs, it outputs the forecast results of various meteorological elements, and evaluates and determines the drought and flood risk level based on the forecast results of various meteorological elements and the preset drought and flood risk level analysis algorithm; According to the correspondence between the drought and flood risk level and the smart home device linkage plan, the smart home device linkage plan is confirmed and implemented.

2. The high temporal and spatial resolution drought and flood conversion prediction and early warning method based on WRF according to claim 1 is characterized in that: According to the correspondence between the regional category and the model parameter setting scheme, the model parameter setting scheme for the area to be studied is analyzed and determined, and the model parameter setting scheme for the area to be studied is implemented, including: Analyze whether there are extreme meteorological events in the study area within the preset time period; If yes, then obtain the data of historical extreme meteorological events, classify the extreme meteorological events according to the data of historical extreme meteorological events, and extract the characteristic parameters of the events; Input the feature parameters of the extracted events into the trained model parameter setting matching model, and output the model parameter setting scheme; If not, keep the original setting.

3. The high temporal and spatial resolution drought-flood conversion prediction and early warning method based on WRF according to claim 1 is characterized by: According to the correspondence between the regional category and the model parameter setting scheme, the model parameter setting scheme for the area to be studied is analyzed and determined, and the model parameter setting scheme for the area to be studied is implemented, including: Obtain change information of the area to be studied, including comprehensive land use change value, comprehensive infrastructure construction progress value, and ecological environment change value; Analyze whether various values ​​of the change information exceed the preset change threshold of the corresponding category; If yes, the regional variation characteristics are used as input objects, input into the relationship model between the trained physical process parameterization scheme and the regional variation characteristics, and output the adjustment parameters of the physical process parameterization scheme; If not, the model parameter setting scheme for the area to be studied is analyzed and determined according to the correspondence between the region category and the model parameter setting scheme, and the model parameter setting scheme for the area to be studied is executed.

4. The high temporal and spatial resolution drought and flood conversion prediction and early warning method based on WRF according to claim 3 is characterized in that: If the change information is the land use change value, the change information of the area to be studied includes: Use satellite remote sensing images to obtain land use data of different expected study areas, and use image classification algorithms to divide the land into different use types, including cultivated land, forest land, construction land, and water areas; Compare and analyze the land use classification results in different periods, and calculate the change area ratio and change speed of each land use type; The two indicators of change area ratio and change speed are standardized, and according to the importance weights of different indicators, the standardized change area ratio and change speed are weighted summed to obtain the comprehensive change value of each land use type; The comprehensive change values ​​of all land use types are summarized and summed to obtain the comprehensive value of land use change.

5. The high temporal and spatial resolution drought-flood conversion prediction and early warning method based on WRF according to claim 3 is characterized by: If the change information is the progress value of comprehensive infrastructure construction, the change information of the area to be studied includes: Collect and obtain relevant data on infrastructure construction in the area to be studied, including the length of road construction, building area, and number of public facilities; According to the corresponding relationship between different infrastructure construction categories and key indicators, determine the key indicators and calculate the progress value of infrastructure construction; The progress value of each infrastructure construction is standardized, and the standardized single progress value is indexed to obtain the single progress index; The weighted sum of each individual progress index is taken to obtain the comprehensive infrastructure construction progress index, and then the comprehensive infrastructure construction progress value is obtained through inverse index transformation.

6. The high temporal and spatial resolution drought and flood conversion prediction and early warning method based on WRF according to claim 3 is characterized in that: If the change information is the ecological environment change value, the change information of the area to be studied includes: Collect and obtain ecological environment-related data, including changes in vegetation coverage, water quality indicators, soil quality indicators, and air quality indicators; For each ecological environment indicator, calculate the change value between different time points, multiply the change value of each indicator by its corresponding weight, and get the weighted change value. Multiply the change value of the indicator by the preset weight to get the weighted change value as the ecological environment change value.

7. The high temporal and spatial resolution drought-flood conversion prediction and early warning method based on WRF according to any one of claims 1 to 6, characterized in that: The preset drought and flood risk level analysis algorithm is as follows: Analyze and extract comprehensive risk index related indicators from the forecast results of various meteorological elements, geographic information data and other data. The comprehensive risk index related indicators include precipitation anomaly percentage, soil moisture index, river water level index and vegetation index change rate; According to the corresponding relationship between the indicators and weights of different comprehensive risk indices, the weight of each indicator is analyzed and determined; According to the weight of each indicator and the related indicators of the comprehensive risk index, the comprehensive risk index calculation formula is used to calculate the comprehensive risk index, which is as follows: ； Among them, the precipitation anomaly percentage weight is , the precipitation anomaly percentage is , the soil moisture index weight is , is the soil moisture index, and the river water level index weight is , is the river water level index, and the vegetation index change rate is , the vegetation index change rate weight is , is the comprehensive risk index; The drought or flood risk level is analyzed and determined based on the correspondence between the range in which the comprehensive risk index falls and the drought or flood risk level.

8. A high temporal and spatial resolution drought and flood conversion prediction and warning system based on WRF, characterized in that: The invention comprises a memory, a processor and a program stored in the memory and executable on the processor, wherein the program can be loaded and executed by the processor to implement a high spatiotemporal resolution drought-flood conversion prediction and early warning method based on WRF as described in any one of claims 1 to 7.

Images