Flash flood forecasting and warning method and device based on refined rainfall analysis

Abstract

The present invention discloses a flash flood forecasting and warning method and device based on refined rainfall analysis. The method includes the following steps: Step 1: Conducting a flash flood survey and evaluation to determine the critical rainfall for each flash flood risk zone; Step 2: Real-time compilation of rainfall data monitored by rainfall gauge stations and spatial interpolation; Step 3: Multi-source data integration and correction processing for short-term and medium-term rainfall forecasts; Step 4: Real-time rolling and refined splicing of the spatially interpolated monitored rainfall, the integrated and corrected short-term and medium-term rainfall forecasts in a temporal dimension; Step 5: Real-time rolling analysis of flash flood disaster risks based on refined rainfall data, and advanced warning. The present invention has the advantages of using highly efficient and accurate refined rainfall data to advance flash flood disaster risk analysis and early warning, reserving valuable time for evacuation and mitigation, and reducing loss of life and property.

Description

Technical Field

[0001] The present invention relates to the technical field of flood and drought disaster prevention, and more specifically, to a flash flood forecasting and early warning method and device based on refined rainfall analysis. Background Art

[0002] Flash floods are caused by heavy rainfall in small watersheds in mountainous areas, resulting in rapid flooding and the subsequent landslides, collapses, and debris flows. They are characterized by rapid onset and severe destructiveness. Therefore, research on flash flood forecasting and early warning is crucial to effectively identify future flash flood disaster risks and provide advance warnings, thereby preserving valuable time for evacuation and mitigation, and thus reducing loss of life and property.

[0003] Flash flood forecasting and warning methods include two approaches: rainfall analysis based on predicted rainfall; and water level analysis based on hydrological and hydrodynamic models. However, the complex runoff generation and convergence mechanisms of flash flood watersheds make suitable hydrological modeling algorithms rare. Flash flood gullies have complex topography and often undergo significant changes under the impact of high-frequency floods. Collecting and updating topographic data to drive hydrodynamic model calculations is expensive. Flash flood gullies, with their rugged terrain, narrow riverbeds, and rapid currents, do not offer suitable conditions for installing hydrological monitoring equipment. Furthermore, the compilation of hydrological data (especially flow data) requires specialized technical personnel, making it extremely difficult to obtain sufficient, high-quality hydrological data to support the calibration and validation of hydrological and hydrodynamic models. Given the limited coverage of flash flood risk areas in my country, flash flood forecasting and warning based on hydrological and hydrodynamic models is extremely difficult to implement on a large scale. On the other hand, the rainfall forecast analysis, forecast and warning method is not effective due to the lack of detailed rainfall data. my country has accelerated the construction of a rainfall monitoring and reporting system with rain gauges and meteorological radars as the core. On this basis, further refined processing of rainfall data and conducting flash flood forecast and warning analysis are important development directions for flash flood disaster prevention. Summary of the Invention

[0004] The purpose of the present invention is to provide a flash flood forecasting and warning method based on refined rainfall analysis, which performs refined processing on monitored, short-term forecast, and medium- and short-term forecast rainfall, and conducts flash flood forecasting and warning analysis in combination with rainfall warning indicators. It can identify future risks and provide advance warnings, improve the warning analysis effect, and reserve time for risk avoidance and evacuation.

[0005] To achieve the above objectives, the present invention provides a first aspect of a flash flood forecasting and warning method based on refined rainfall analysis, comprising:

[0006] Collect and process data on flash flood risk areas, determine the critical rainfall for each flash flood risk area, and obtain the rainfall warning index for flash flood risk based on the critical rainfall for each flash flood risk area;

[0007] Perform real-time compilation and spatial interpolation of rainfall data monitored by rainfall stations to obtain spatially interpolated rainfall data;

[0008] Integrate and revise multi-source short-term and short-term rainfall forecast data and multi-source medium- and short-term rainfall forecast data respectively;

[0009] The spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised medium- and short-term forecast rainfall data are finely spliced in the time dimension to obtain refined rainfall data;

[0010] Analysis and early warning of flash flood disasters based on refined rainfall data and rainfall warning indicators of flash flood risks.

[0011] In one embodiment, data collection and processing are performed on flash flood risk areas to determine the critical rainfall for each flash flood risk area, including:

[0012] S11: Investigate and collect historical flood data of control sections in flash flood risk areas and measure the topography of the sections;

[0013] S12: Extract the rainwater collection area of the control section based on the digital elevation model;

[0014] S13: Calculate the design rainstorm in the catchment area of the control section, and obtain the design flood process of the control section through the calculation of runoff generation, confluence, and water level-discharge relationship;

[0015] S14: Based on the design flood process of the control section, the critical water level is determined using the historical flood analysis method, and step S13 is repeated and repeated calculations are performed assuming the initial rainfall to obtain the critical water level. The rainfall corresponding to the critical water level is the critical rainfall;

[0016] S15: Based on the critical rainfall, the rainfall for ready transfer and the rainfall for immediate transfer in different periods are obtained as rainfall warning indicators for flash flood risks.

[0017] In one embodiment, real-time compilation and spatial interpolation of rainfall data monitored by rainfall gauges are performed to obtain spatially interpolated rainfall data, including:

[0018] During rainfall periods, the frequency of collecting monitoring data at rainfall stations will be increased to the minute level, and the data will be updated in real time to the rainfall during the hour. For rainfall monitored across the hour, the rainfall will be allocated to different time periods according to the proportion of time length.

[0019] The hourly monitoring rainfall data of the rain gauge station is interpolated to each grid in the region based on the spatial interpolation technology. The hourly monitoring rainfall data of the rain gauge station includes the historical hourly rainfall data and the hourly rainfall data that is updated in real time during the hour.

[0020] In one embodiment, integrating and correcting multi-source short-term and short-term rainfall forecast data and multi-source medium-term and short-term rainfall forecast data respectively includes:

[0021] Real-time connection to multi-source short-term and short-term rainfall forecast data, and multi-source medium-term and short-term rainfall forecast data;

[0022] Resample the short-term and short-term rainfall forecast data from multiple sources and the medium-term and short-term rainfall forecast data from multiple sources, calculate the indicators used to evaluate the data accuracy of the rainfall forecast products from each source, and calculate the weight coefficients for the integration of the rainfall forecast products from each source based on the calculated indicators;

[0023] According to the weight coefficient of the integration of each source's forecast rainfall product, the multi-source short-term forecast rainfall data and the multi-source medium-term forecast rainfall data are integrated respectively;

[0024] The integrated short-term forecast rainfall data and the integrated medium- and short-term forecast rainfall data are corrected.

[0025] In one embodiment, the correction processing of the integrated short-term and short-term rainfall forecast data and the integrated medium-term and short-term rainfall forecast data includes using a frequency matching method to perform correction processing, specifically:

[0026] Sampling is performed based on a time sliding window, and the sample collection content includes the monitored rainfall and integrated forecast rainfall at the rain gauge location;

[0027] Collected sample data, statistically calculate a set of rainfall thresholds Corresponding integrated forecast rainfall accumulation frequency and monitor rainfall accumulation frequency ,in, is the number of sample rainfall data, is the Nth rainfall threshold, for The corresponding rainfall accumulation frequency, for The corresponding monitoring rainfall accumulation frequency;

[0028] Based on linear interpolation or quadratic sample interpolation, respectively, through the coordinate ,coordinate Construct integrated forecast rainfall accumulation frequency curve and monitoring rainfall accumulation frequency curve;

[0029] For the forecast rainfall value of any grid in the integrated short-term forecast rainfall data or the integrated medium-term forecast rainfall data, the cumulative frequency is obtained through the forecast rainfall cumulative frequency curve, and the monitored rainfall value corresponding to the cumulative frequency is obtained according to the monitored rainfall cumulative frequency curve. The forecast rainfall value of the grid is then corrected to the obtained monitored rainfall value, and finally the integrated and corrected short-term forecast rainfall data or the integrated and corrected medium-term forecast rainfall data are obtained.

[0030] In one embodiment, the spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised short-term forecast rainfall data are finely spliced in the time dimension, including:

[0031] Resample the spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised short-term forecast rainfall data to the same spatial resolution;

[0032] The resampled short-term forecast rainfall data is compiled into hourly rainfall data, including the short-term forecast rainfall data for the remaining period of the current hour, the entire period of the next hour, and part of the period of the next hour;

[0033] According to the ratio of the remaining period of the next hour to the complete period of the next hour, the rainfall of the next hour of the medium- and short-term forecast rainfall data is distributed to obtain the medium- and short-term forecast rainfall data of the remaining period of the next hour;

[0034] The rainfall monitoring data of historical periods and part of the current hour with the same spatial resolution, the integrated and revised short-term forecast rainfall data of the remaining period of the current hour, the next hour, and part of the next hour, and the integrated and revised medium- and short-term forecast rainfall data of the remaining period of the next hour and thereafter are finely spliced in the time dimension to obtain refined period rainfall data covering historical long-series monitoring to the future 72-h forecast.

[0035] In one embodiment, flash flood disaster analysis and early warning based on refined rainfall data and flash flood risk rainfall warning indicators include:

[0036] Extracting rainfall data of a first preset period in the past and rainfall data of a second preset period in the future from the refined rainfall data;

[0037] Based on the intercepted refined rainfall data, the surface period rainfall of the rainwater collection area in each flash flood risk zone from the first preset period in the past to the second preset period in the future is calculated;

[0038] Based on the surface period rainfall from the first preset period in the past to the second preset period in the future, combined with the prepared transfer rainfall and immediate transfer rainfall in different periods, a real-time rolling analysis of the flash flood disaster risk in the future period is conducted;

[0039] Use modern communication technology to send targeted information on flash flood disaster risks in the future.

[0040] Based on the same inventive concept, the second aspect of the present invention provides a flash flood forecasting and warning device based on refined rainfall analysis, comprising:

[0041] The data collection and processing module is used to collect and process data from flash flood risk areas, determine the critical rainfall of each flash flood risk area, and obtain the rainfall warning index of flash flood risk based on the critical rainfall of the flash flood risk area;

[0042] The monitoring rainfall compilation and interpolation module is used to perform real-time compilation and spatial interpolation of rainfall data monitored by rainfall stations to obtain spatially interpolated monitoring rainfall data;

[0043] Multi-source forecast rainfall data integration and correction module, used to integrate and correct multi-source short-term forecast rainfall data and multi-source medium-term forecast rainfall data respectively;

[0044] The data splicing module is used to finely splice the spatially interpolated monitoring rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised medium- and short-term forecast rainfall data in the time dimension to obtain refined rainfall data;

[0045] Analysis and early warning of flash flood disasters based on refined rainfall data and rainfall warning indicators of flash flood risk

[0046] Based on the same inventive concept, the third aspect of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the flash flood forecasting and warning method based on refined rainfall analysis described in the first aspect.

[0047] Based on the same inventive concept, the fourth aspect of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the program, the flash flood forecasting and warning method based on refined rainfall analysis described in the first aspect is implemented.

[0048] Compared with the prior art, the advantages and beneficial technical effects of the present invention are as follows:

[0049] The present invention provides a flash flood forecasting and warning method based on refined rainfall analysis. First, data of flash flood risk areas are collected and processed to determine the critical rainfall of each flash flood risk area, and further rainfall warning indicators of flash flood risks are obtained. Multi-source short-term forecast rainfall data and multi-source medium-term forecast rainfall data are integrated and corrected respectively, which can improve the stability and numerical accuracy of short-term and medium-term forecast rainfall. Rainfall station monitoring data are real-time reorganized and spatially interpolated, and the interpolated rasterized monitoring rainfall, integrated-corrected short-term forecast rainfall, and integrated-corrected medium-term forecast rainfall are finely spliced in the time dimension, ensuring the high timeliness and high accuracy of rainfall data over the entire time period. Based on the refined forecast rainfall and the rainfall warning indicators for flash flood risks, the flash flood disaster risk in the next 24 hours is analyzed and advanced warning is carried out, which improves the warning effect, can reserve precious time for risk avoidance and evacuation of flash flood disasters, and reduce loss of life and property. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 Flowchart of a flash flood forecasting and warning method based on refined rainfall analysis disclosed in an embodiment of the present invention;

[0052] Figure 2 A comparison chart of the integrated-corrected forecast rainfall and the monitored rainfall in an embodiment of the present invention;

[0053] Figure 3 1 is a schematic diagram of a flash flood forecast and warning device based on refined rainfall analysis in an embodiment of the present invention;

[0054] Figure 4 A schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of the present invention;

[0055] Figure 5 A schematic diagram of the structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The present invention discloses a flash flood forecasting and early warning method based on refined rainfall analysis, comprising the following steps: step one: conducting a flash flood investigation and evaluation, determining the critical rainfall of each flash flood risk area, and obtaining a flash flood risk rainfall early warning index based on the critical rainfall of the flash flood risk area; step two: real-time compilation of rainfall data monitored by rainfall stations and spatial interpolation; step three: multi-source data integration and correction processing of short-term forecast and medium-term forecast rainfall; step four: real-time rolling and refined splicing of spatially interpolated monitored rainfall, integrated-corrected short-term forecast rainfall, and integrated-corrected medium-term forecast rainfall in a time dimension; step five: real-time rolling analysis of flash flood disaster risks based on the refined rainfall data and the flash flood risk rainfall early warning index, and providing advance warnings. The present invention has the advantages of using high-efficiency and high-precision refined rainfall to analyze flash flood disaster risks in advance and issue early warnings, which can reserve valuable time for risk avoidance and evacuation, and reduce loss of life and property (high timeliness is reflected in the refined splicing of three types of rainfall data in the time dimension; high precision is reflected in the integration and correction processing of multi-source data to improve rainfall accuracy).

[0057] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0058] Example 1

[0059] This embodiment discloses a flash flood forecasting and warning method based on refined rainfall analysis. Figure 1 ,include:

[0060] S1: Collect and process data on flash flood risk areas, determine the critical rainfall of each flash flood risk area, and obtain the rainfall warning index of flash flood risk based on the critical rainfall of the flash flood risk area;

[0061] S2: Real-time compilation and spatial interpolation of rainfall data monitored by rainfall stations to obtain spatially interpolated rainfall data;

[0062] S3: Integrate and correct multi-source short-term and short-term rainfall forecast data and multi-source medium-term and short-term rainfall forecast data respectively;

[0063] S4: The spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised medium- and short-term forecast rainfall data are finely spliced in the time dimension to obtain refined rainfall data;

[0064] S5: Analysis and early warning of flash flood disasters based on refined rainfall data and rainfall warning indicators of flash flood risks.

[0065] Specifically, through analysis of existing technologies, the applicant found that the existing flash flood warning methods mainly have the following two problems: 1) The water level analysis, forecast and warning method is restricted by many factors and is difficult to promote and apply on a large scale; 2) The rainfall analysis, forecast and warning method is not effective due to the lack of detailed rainfall data. At present, my country has accelerated the construction of a rainfall measurement and reporting system, and the fine-tuning of rainfall data on this basis can improve the warning effect.

[0066] The spatial interpolation method in S2 can use methods such as inverse distance weighting method and kriging method. Short-term forecast rainfall data generally refers to rainfall data with a forecast period of 2 hours, and medium-term forecast rainfall data generally refers to rainfall data with a forecast period of 3 days or more.

[0067] In one embodiment, data collection and processing are performed on flash flood risk areas to determine the critical rainfall for each flash flood risk area, including:

[0068] S11: Investigate and collect historical flood data of control sections in flash flood risk areas and measure the topography of the sections;

[0069] S12: Extract the rainwater collection area of the control section based on the digital elevation model;

[0070] S13: Calculate the design rainstorm in the catchment area of the control section, and obtain the design flood process of the control section through the calculation of runoff (net rain), confluence, and water level-discharge relationship;

[0071] S14: Based on the design flood process of the control section, the critical water level is determined using the historical flood analysis method, and step S13 is repeated and repeated calculations are performed assuming the initial rainfall to obtain the critical water level. The rainfall corresponding to the critical water level is the critical rainfall;

[0072] S15: Based on the critical rainfall, the rainfall for ready transfer and the rainfall for immediate transfer in different periods are obtained as rainfall warning indicators for flash flood risks.

[0073] Specifically, S12 performs a series of operations on the digital elevation model, including digital river channel correction, depression filling, flow direction generation, and grid catchment area calculation, to extract the catchment area of the control section. The flood process calculated in S13 includes water level and flow processes. Critical rainfall in S14 includes 1-hour critical rainfall, 3-hour critical rainfall, and 6-hour critical rainfall.

[0074] In one embodiment, real-time compilation and spatial interpolation of rainfall data monitored by rainfall gauges are performed to obtain spatially interpolated rainfall data, including:

[0075] During rainfall periods, the frequency of collecting monitoring data at rainfall stations will be increased to the minute level, and the data will be updated in real time to the rainfall during the hour. For rainfall monitored across the hour, the rainfall will be allocated to different time periods according to the proportion of time length.

[0076] The hourly monitoring rainfall data of the rain gauge station is interpolated to each grid in the region based on the spatial interpolation technology. The hourly monitoring rainfall data of the rain gauge station includes the historical hourly rainfall data and the hourly rainfall data that is updated in real time during the hour.

[0077] During the specific implementation process, when the rainfall data monitored by the rain gauge station is compiled in real time, the rainfall from 11:15 to 11:20 is collected at 11:20, then the rainfall from 11:00 to 12:00 should be updated to the sum of the rainfall from 11:00 to 11:15 and the rainfall from 11:15 to 11:20, and so on, until all time periods of the hour are covered.

[0078] For the monitored rainfall across the hour, the rainfall is allocated to different time periods according to the ratio of time length. For example, if the rainfall from 11:52 to 12:05 is 10 mm, then 8 / 13×10 mm (i.e. 6.2 mm) is allocated to 11:00-12:00, and 5 / 1310 mm (i.e. 3.8 mm) is allocated to 12:00-13:00.

[0079] When performing spatial interpolation, the grid resolution depends on the overall situation of the catchment area of all flash flood risk areas. Generally, a resolution of 1km×1km is recommended.

[0080] In one embodiment, integrating and correcting multi-source short-term and short-term rainfall forecast data and multi-source medium-term and short-term rainfall forecast data respectively includes:

[0081] Real-time connection to multi-source short-term and short-term rainfall forecast data, and multi-source medium-term and short-term rainfall forecast data;

[0082] Resample the short-term and short-term rainfall forecast data from multiple sources and the medium-term and short-term rainfall forecast data from multiple sources, calculate the indicators used to evaluate the data accuracy of the rainfall forecast products from each source, and calculate the weight coefficients for the integration of the rainfall forecast products from each source based on the calculated indicators;

[0083] According to the weight coefficient of the integration of each source's forecast rainfall product, the multi-source short-term forecast rainfall data and the multi-source medium-term forecast rainfall data are integrated respectively;

[0084] The integrated short-term forecast rainfall data and the integrated medium- and short-term forecast rainfall data are corrected.

[0085] Specifically, multi-source short-term and short-term rainfall forecast data, or multi-source short-term and medium-term rainfall forecast data, are referred to as multi-source short-term and short-term rainfall forecast products. Generally, multi-source short-term rainfall forecast data (with a forecast period of 2 hours) has consistent temporal and spatial resolutions, and update frequencies across all sources: 5 minutes, 1 km × 1 km, and updated every 5 minutes, respectively.

[0086] As for multi-source medium- and short-term rainfall forecasts (forecast period of 3 days or longer), such as the ECMWF produced by the European Centre for Medium-Range Weather Forecasts (the time resolution for the first 3 days is 1 hour, the spatial resolution is 9 km × 9 km, and it is updated every 12 hours), the GFS produced by the National Oceanic and Atmospheric Administration of the United States (the time resolution for the first 3 days is 1 hour, the spatial resolution is 25 km × 25 km, and it is updated every 6 hours), and the CMA produced by the China Meteorological Administration (the forecast period is 3 days, the time resolution is 1 hour, the spatial resolution is 2 km × 2 km, and it is updated every 12 hours).

[0087] The short-term and medium-term multi-source rainfall forecast data are integrated and processed to improve the stability of rainfall forecast and reduce the instability of a certain type of forecast rainfall data under specific scenarios. The integrated processing of short-term and medium-term multi-source rainfall forecast data includes the following steps:

[0088] (1) Resample the forecast rainfall data (short-term, medium-term and short-term multi-source forecast rainfall data) to keep the spatial resolution of each source data consistent;

[0089] (2) Sampling is performed based on a time sliding window. The window range is one month before and after the same period of previous years, and the window step is 1 day. The sample collection content includes the monitored rainfall at the rain gauge location and the multi-source forecast rainfall.

[0090] (3) Using the sample data, evaluate the data accuracy of rainfall forecast products from each source based on continuity indicators (such as correlation coefficient CC, mean error ME, root mean square error RMSE, etc.) or classification indicators (such as TS score, detection rate POD, false alarm rate FAR, etc.), and use them as the basis for integration. The RMSE calculation formula is:

[0091]

[0092] Where, To forecast rainfall, monitoring rainfall for rain gauges; is the sample size.

[0093] (4) Calculate the weight coefficient of the integration of each source forecast rainfall product, and integrate the multi-source forecast rainfall data connected in real time into the new forecast rainfall data. If the multi-source forecast rainfall data is integrated based on RMSE, the weight coefficient of the kth forecast rainfall product integration is Use the following formula to calculate:

[0094]

[0095] The rainfall P of the integrated forecast rainfall product is:

[0096]

[0097] Where, Rainfall data for forecast rainfall products from various sources.

[0098] In one embodiment, the correction processing of the integrated short-term and short-term rainfall forecast data and the integrated medium-term and short-term rainfall forecast data includes using a frequency matching method to perform correction processing, specifically:

[0099] Sampling is performed based on a time sliding window, and the sample collection content includes the monitored rainfall and integrated forecast rainfall at the rain gauge location;

[0100] Collected sample data, statistically calculate a set of rainfall thresholds Corresponding integrated forecast rainfall accumulation frequency and monitor rainfall accumulation frequency ,in, is the number of sample rainfall data, is the Nth rainfall threshold, for The corresponding rainfall accumulation frequency, for The corresponding monitoring rainfall accumulation frequency;

[0101] Based on linear interpolation or quadratic sample interpolation, respectively, through the coordinate ,coordinate Construct integrated forecast rainfall accumulation frequency curve and monitoring rainfall accumulation frequency curve;

[0102] For the forecast rainfall value of any grid in the integrated short-term forecast rainfall data or the integrated medium-term forecast rainfall data, the cumulative frequency is obtained through the forecast rainfall cumulative frequency curve, and the monitored rainfall value corresponding to the cumulative frequency is obtained according to the monitored rainfall cumulative frequency curve. The forecast rainfall value of the grid is then corrected to the obtained monitored rainfall value, and finally the integrated and corrected short-term forecast rainfall data or the integrated and corrected medium-term forecast rainfall data are obtained.

[0103] Specifically, real-time rolling corrections are performed on the integrated short-term and medium-term rainfall forecasts to reduce forecast errors and improve forecast accuracy. This real-time rolling correction employs a frequency matching method, which uses deterministic forecasts from the training period and corresponding actual data (i.e., rainfall measured at rain gauges) to establish a one-to-one mapping between the original forecast and the forecast correction. This method aims to eliminate frequency bias across different precipitation levels, ensuring consistency between the total area of the forecast and the observed rainfall area. Sampling is based on a sliding time window, spanning one month before and after the same period in previous years, with a step size of one hour.

[0104] In other implementations, other methods may be used for correction processing, such as geostatistical analysis, LSTM method, etc.

[0105] During implementation, the upper and lower bounds of the correction were set at 50 and 0.1 mm / h, respectively, to prevent excessive corrections due to extreme precipitation extrapolation and the disappearance of light rain. If the original forecast exceeded the upper bound, the original value was retained; if the revised forecast fell below the lower bound, the rainfall was set to zero.

[0106] In one embodiment, the spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised short-term forecast rainfall data are finely spliced in the time dimension, including:

[0107] Resample the spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised short-term forecast rainfall data to the same spatial resolution;

[0108] The resampled short-term forecast rainfall data is compiled into hourly rainfall data, including the short-term forecast rainfall data for the remaining period of the current hour, the entire period of the next hour, and part of the period of the next hour;

[0109] According to the ratio of the remaining period of the next hour to the complete period of the next hour, the rainfall of the next hour of the medium- and short-term forecast rainfall data is distributed to obtain the medium- and short-term forecast rainfall data of the remaining period of the next hour;

[0110] The rainfall monitoring data of historical periods and part of the current hour with the same spatial resolution, the integrated and revised short-term forecast rainfall data of the remaining period of the current hour, the next hour, and part of the next hour, and the integrated and revised medium- and short-term forecast rainfall data of the remaining period of the next hour and thereafter are finely spliced in the time dimension to obtain refined period rainfall data covering historical long-series monitoring to the future 72-h forecast.

[0111] Specifically, the spatial resolution of the resampling can be set to 1km×1km.

[0112] The remaining time of this hour refers to the time period after the current time, and the partial time period of this hour refers to the time period before the current time.

[0113] In one embodiment, flash flood disaster analysis and early warning based on refined rainfall data and flash flood risk rainfall warning indicators include:

[0114] Extracting rainfall data of a first preset period in the past and rainfall data of a second preset period in the future from the refined rainfall data;

[0115] Based on the intercepted refined rainfall data, the surface period rainfall of the rainwater collection area in each flash flood risk zone from the first preset period in the past to the second preset period in the future is calculated;

[0116] Based on the surface period rainfall from the first preset period in the past to the second preset period in the future, combined with the prepared transfer rainfall and immediate transfer rainfall in different periods, a real-time rolling analysis of the flash flood disaster risk in the future period is conducted;

[0117] Use modern communication technology to send targeted information on flash flood disaster risks in the future.

[0118] Specifically, given that the time periods for flash flood rainfall warning analysis are generally 1 hour, 3 hours, and 6 hours, the first preset time period is set to 5 hours, that is, the rainfall data for the past 5 hours (flash flood rainfall warning analysis) is intercepted from the refined rainfall data; given that the error of the rainfall forecast for the next 24 hours is relatively small, the second preset time period is set to 24 hours, that is, the rainfall data for the next 24 hours is intercepted from the refined rainfall data;

[0119] In this embodiment, the surface period rainfall from the first preset period in the past to the second preset period in the future represents the surface period rainfall from the past 5 hours to the future 24 hours. Based on the refined rasterized rainfall data, the surface period rainfall for the rainfall collection range of each flash flood risk area (past 5 hours - future 24 hours) is calculated. If flash flood ditch vector data or flash flood risk area rainfall collection range vector data cannot be collected, it is necessary to find the period rainfall data for the corresponding grid of the flash flood risk area (or its associated rain gauge station);

[0120] Based on detailed rainfall data (either surface data from the catchment area or grid data for flash flood risk areas) for the past 5 hours to the next 24 hours, combined with the prepared and immediate rainfall for different time periods (1 hour, 3 hours, and 6 hours), a real-time rolling analysis of flash flood disaster risk over the next 24 hours is conducted, including risk level and time of occurrence. Flash flood disaster analysis results based on different time periods may differ at the same moment; in these cases, the maximum possible risk is determined.

[0121] By using modern communication technologies (such as SMS services, LBS technology (Location Based Services), virtual electronic fences, remote shouting, etc.), flash flood disaster risk information for the next 24 hours will be sent to those responsible for flash floods, reserving time for evacuation.

[0122] The present invention is now described in detail by taking the flash flood early warning analysis of a certain district and county in Chongqing as an example, which also has a guiding role in the application of the present invention to flash flood monitoring and early warning in other areas.

[0123] Refer to the attached figure: Figure 1 As shown, in this embodiment, the flash flood forecasting and warning method based on refined rainfall analysis includes the following steps:

[0124] (1) Carry out flash flood dispatch evaluation in more than 100 flash flood risk areas. After extracting the rainwater collection range, calculating the design rainstorm, design flood (runoff generation), and water level-discharge relationship, the design flood process of the control section is obtained. Based on the historical flood analysis, repeated trial calculations are carried out to analyze and determine the ready-to-transfer rainfall and immediate transfer rainfall in different time periods (1 hour, 3 hours, and 6 hours) as rainfall warning indicators for flash flood risks.

[0125] (2) During the rainfall period, the frequency of collecting monitoring data at the rain gauge station is increased to the minute level, and the data is updated in real time to the rainfall of the hour. For the monitored rainfall across the hour, the rainfall can be allocated to different periods according to the ratio of time length. Based on the inverse distance weighted method, the hourly monitored rainfall of the rain gauge station (including the historical hourly rainfall and the hourly rainfall updated in real time) is interpolated to each grid in the region, with a grid resolution of 1 km × 1 km.

[0126] (3) In real time, the Caiyun short-term forecast rainfall data (with a time resolution of 5 minutes, a spatial resolution of 1 km × 1 km, and updated every 5 minutes) are connected with the medium-term and short-term forecast rainfall data from multiple sources (including CMA and ECMWF), and the ECMWF with a resolution of 9 km × 9 km is resampled to 9 km × 9 km. Based on a time sliding window covering a range of one month before and after the same period of previous years and a step length of 1 hour, the monitored rainfall at the rain gauge location, the Caiyun short-term forecast rainfall, the CMA medium-term and short-term forecast rainfall, and the ECMWF medium-term and short-term forecast rainfall sample data are obtained. Based on the RMSE evaluation results of the forecast rainfall products from each source, the CMA and ECMWF forecast rainfall data are integrated and processed. Based on the frequency matching method, the Caiyun short-term forecast rainfall and the integrated medium-term and short-term forecast rainfall are subjected to real-time rolling correction processing to obtain the integrated-corrected short-term forecast rainfall and the integrated-corrected medium-term and short-term forecast rainfall (see Figure 2 );

[0127] (4) Resample the integrated and revised short-term forecast rainfall data to a resolution of 1 km × 1 km; compile the short-term forecast rainfall data (next 2 hours) with a time resolution of 5 minutes into hourly rainfall data, including the short-term forecast rainfall data for the remaining period of the current hour (after the current moment), the entire period of the next hour, and the partial period of the next hour; allocate the rainfall for the next hour of the medium-term forecast according to the ratio of the duration of the remaining period of the next hour to the complete period, and obtain the short-term forecast rainfall data for the remaining period of the next hour; finely splice the rainfall monitoring data of the historical period and the partial period of the current hour (before the current moment) with the same spatial resolution, the integrated and revised short-term forecast rainfall data for the remaining period of the current hour (after the current moment), the next hour, and the partial period of the next hour, and the integrated and revised short-term forecast rainfall data for the remaining period of the next hour and thereafter hourly in the time dimension to obtain the refined period rainfall data covering the "historical long-series monitoring-next 72-hour forecast", ensuring that the rainfall data has extremely high timeliness and numerical accuracy.

[0128] (5) Extract rainfall data for the past 5 hours and the next 24 hours from the refined rainfall forecast, and find the rainfall data for the corresponding grid in the flash flood risk area; Based on the refined rainfall data for the period of "past 5 hours - next 24 hours", combined with the ready transfer rainfall and immediate transfer rainfall in different periods (1 hour, 3 hours, 6 hours), conduct a real-time rolling analysis of the flash flood disaster risk in the next 24 hours, including the risk level and the time of risk occurrence; Use modern communication technology (including SMS service, LBS technology, virtual electronic fence, remote shouting, etc.) to send the flash flood disaster risk information for the next 24 hours to the flash flood responsible person, and reserve transfer time for flash flood risk avoidance.

[0129] Conclusion: This embodiment adopts the method of the present invention to analyze the flash flood disaster risk in the next 24 hours with high timeliness and high precision and provide advance warning, which can reserve valuable time for risk avoidance and evacuation, reduce loss of life and property, and has significant social and economic benefits.

[0130] Example 2

[0131] Based on the same inventive concept, this embodiment discloses a flash flood forecasting and warning device based on refined rainfall analysis, see Figure 3 ,include:

[0132] The data collection and processing module 101 is used to collect and process data on flash flood risk areas, determine the critical rainfall of each flash flood risk area, and obtain a flash flood risk rainfall warning indicator based on the critical rainfall of the flash flood risk area;

[0133] The monitoring rainfall compilation and interpolation module 102 is used to perform real-time compilation and spatial interpolation of the rainfall data monitored by the rainfall station to obtain spatially interpolated monitoring rainfall data;

[0134] The multi-source rainfall forecast data integration and correction module 103 is used to integrate and correct the multi-source short-term rainfall forecast data and the multi-source medium-term and short-term rainfall forecast data respectively;

[0135] The data splicing module 104 is used to perform fine splicing of the spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised medium-term forecast rainfall data in the time dimension to obtain refined rainfall data;

[0136] The analysis and warning module 105 is used to analyze and warn flash flood disasters based on refined rainfall data and rainfall warning indicators of flash flood risks. In one embodiment, the device also includes an artificial intelligence security and image credibility module 205, which is used to perform content review on the enhanced image, evaluate its authenticity and credibility by analyzing the feature distribution visibility, detail rationality, and distortion of the enhanced image, and generate image data with anti-counterfeiting characteristics by embedding digital watermarks and performing image content monitoring.

[0137] Since the device described in Example 2 of the present invention is used to implement the flash flood forecast and warning method based on refined rainfall analysis in Example 1 of the present invention, those skilled in the art will be able to understand the specific structure and variations of the device based on the method described in Example 1 of the present invention, and therefore will not be described in detail here. All devices used in the method of Example 1 of the present invention fall within the scope of protection of the present invention.

[0138] Example 3

[0139] Based on the same inventive concept, see Figure 4 The present invention further provides a computer-readable storage medium 300 on which a computer program 311 is stored. When the program is executed by a processor, the method described in the first embodiment is implemented.

[0140] Since the computer-readable storage medium described in Example 3 of the present invention is the computer-readable storage medium used to implement the flash flood forecast and warning method based on refined rainfall analysis in Example 1 of the present invention, those skilled in the art will be able to understand the specific structure and variations of the computer-readable storage medium based on the method described in Example 1 of the present invention, and therefore will not be described in detail here. All computer-readable storage media used in the method of Example 1 of the present invention fall within the scope of protection of the present invention.

[0141] Example 4

[0142] Based on the same inventive concept, see Figure 5The present invention further provides a computer device, comprising a memory 401, a processor 402, and a computer program 403 stored in the memory and executable on the processor, wherein the processor implements the method described in the first embodiment when executing the program.

[0143] Since the computer device described in Example 4 of the present invention is the computer device used to implement the flash flood forecast and warning method based on refined rainfall analysis in Example 1 of the present invention, those skilled in the art will be able to understand the specific structure and variations of the computer device based on the method described in Example 1 of the present invention, and therefore will not be described in detail here. All computer devices used in the method of Example 1 of the present invention fall within the scope of protection of the present invention.

[0144] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0145] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0146] Although preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they are aware of the basic inventive concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention. Obviously, those skilled in the art may make various changes and modifications to the embodiments of the present invention without departing from the spirit and scope of the embodiments of the present invention. Thus, the present invention is intended to include such changes and modifications to the embodiments of the present invention if they fall within the scope of the claims and their equivalents.

Claims

1. A flash flood forecasting and warning method based on refined rainfall analysis, characterized in that: include: Collect and process data on flash flood risk areas, determine the critical rainfall for each flash flood risk area, and obtain the rainfall warning index for flash flood risk based on the critical rainfall for each flash flood risk area; The rainfall data monitored by the rain gauge station is compiled and spatially interpolated in real time to obtain spatially interpolated monitoring rainfall data. Specifically, during the rainfall period, the collection frequency of the monitoring data of the rain gauge station is increased to the minute level, and it is updated in real time to the rainfall in the current hour. Among them, for the monitoring rainfall across the hour, the rainfall is allocated to different time periods according to the time length ratio; based on the spatial interpolation technology, the hourly monitoring rainfall data of the rain gauge station is interpolated to each grid in the region. Among them, the hourly monitoring rainfall data of the rain gauge station includes the historical compiled hourly rainfall data and the hourly rainfall data compiled and updated in real time in the current hour; Integrate and revise multi-source short-term and short-term rainfall forecast data and multi-source medium- and short-term rainfall forecast data respectively; The spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised medium- and short-term forecast rainfall data are finely spliced in the time dimension to obtain refined rainfall data. Analysis and early warning of flash flood disasters based on refined rainfall data and rainfall warning indicators of flash flood risks.

2. The flash flood forecasting and warning method based on refined rainfall analysis according to claim 1, characterized in that: Collect and process data on flash flood risk areas and determine the critical rainfall for each flash flood risk area, including: S11: Investigate and collect historical flood data of control sections in flash flood risk areas and measure the topography of the sections; S12: Extract the rainwater collection area of the control section based on the digital elevation model; S13: Calculate the design rainstorm in the catchment area of the control section, and obtain the design flood process of the control section through the calculation of runoff generation, confluence, and water level-discharge relationship; S14: Based on the design flood process of the control section, the critical water level is determined using the historical flood analysis method, and step S13 is repeated and repeated calculations are performed assuming the initial rainfall to obtain the critical water level. The rainfall corresponding to the critical water level is the critical rainfall; S15: Based on the critical rainfall, the rainfall for ready transfer and the rainfall for immediate transfer in different periods are obtained as rainfall warning indicators for flash flood risks.

3. The flash flood forecasting and warning method based on refined rainfall analysis according to claim 1, characterized in that: The multi-source short-term and short-term rainfall forecast data and the multi-source medium-term and short-term rainfall forecast data are integrated and corrected respectively, including: Real-time connection to multi-source short-term and short-term rainfall forecast data, and multi-source medium-term and short-term rainfall forecast data; Resample the short-term and short-term rainfall forecast data from multiple sources and the medium-term and short-term rainfall forecast data from multiple sources, calculate the indicators used to evaluate the data accuracy of the rainfall forecast products from each source, and calculate the weight coefficients for the integration of the rainfall forecast products from each source based on the calculated indicators; According to the weight coefficient of the integration of each source's forecast rainfall product, the multi-source short-term forecast rainfall data and the multi-source medium-term forecast rainfall data are integrated respectively; The integrated short-term forecast rainfall data and the integrated medium- and short-term forecast rainfall data are corrected.

4. The flash flood forecasting and warning method based on refined rainfall analysis according to claim 3, characterized in that: Correction processing of the integrated short-term forecast rainfall data and the integrated medium-term forecast rainfall data includes correction processing using the frequency matching method, specifically: Sampling is performed based on a time sliding window, and the sample collection content includes the monitored rainfall and integrated forecast rainfall at the rain gauge location; Collected sample data, statistically calculate a set of rainfall thresholds Corresponding integrated forecast rainfall accumulation frequency and monitor rainfall accumulation frequency ,in, is the number of sample rainfall data, is the Nth rainfall threshold, for The corresponding rainfall accumulation frequency, for The corresponding monitoring rainfall accumulation frequency; Based on linear interpolation or quadratic sample interpolation, respectively, through the coordinate ,coordinate Construct integrated forecast rainfall accumulation frequency curve and monitoring rainfall accumulation frequency curve; For the forecast rainfall value of any grid in the integrated short-term forecast rainfall data or the integrated medium-term forecast rainfall data, the cumulative frequency is obtained through the forecast rainfall cumulative frequency curve, and the monitored rainfall value corresponding to the cumulative frequency is obtained according to the monitored rainfall cumulative frequency curve. The forecast rainfall value of the grid is then corrected to the obtained monitored rainfall value, and finally the integrated and corrected short-term forecast rainfall data or the integrated and corrected medium-term forecast rainfall data are obtained.

5. The flash flood forecasting and warning method based on refined rainfall analysis according to claim 1, characterized in that: The spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised medium- and short-term forecast rainfall data are finely spliced in the time dimension, including: Resample the spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised short-term forecast rainfall data to the same spatial resolution; The resampled short-term forecast rainfall data is compiled into hourly rainfall data, including the short-term forecast rainfall data for the remaining period of the current hour, the entire period of the next hour, and part of the period of the next hour; According to the ratio of the remaining period of the next hour to the complete period of the next hour, the rainfall of the next hour of the medium- and short-term forecast rainfall data is distributed to obtain the medium- and short-term forecast rainfall data of the remaining period of the next hour; The rainfall monitoring data of historical periods and part of the current hour with the same spatial resolution, the integrated and revised short-term forecast rainfall data of the remaining period of the current hour, the next hour, and part of the next hour, and the integrated and revised medium- and short-term forecast rainfall data of the remaining period of the next hour and thereafter are finely spliced in the time dimension to obtain refined period rainfall data covering historical long-series monitoring to the future 72-h forecast.

6. The flash flood forecasting and warning method based on refined rainfall analysis according to claim 2, characterized in that: Flash flood disaster analysis and early warning based on refined rainfall data and flash flood risk warning indicators, including: Extracting rainfall data of a first preset period in the past and rainfall data of a second preset period in the future from the refined rainfall data; Based on the intercepted refined rainfall data, the surface period rainfall of the rainwater collection area in each flash flood risk zone from the first preset period in the past to the second preset period in the future is calculated; Based on the surface period rainfall from the first preset period in the past to the second preset period in the future, combined with the prepared transfer rainfall and immediate transfer rainfall in different periods, a real-time rolling analysis of the flash flood disaster risk in the future period is conducted; Use modern communication technology to send targeted information on flash flood disaster risks in the future.

7. A flash flood forecast and warning device based on refined rainfall analysis, characterized in that: include: The data collection and processing module is used to collect and process data from flash flood risk areas, determine the critical rainfall of each flash flood risk area, and obtain the rainfall warning index of flash flood risk based on the critical rainfall of the flash flood risk area; The monitoring rainfall compilation and interpolation module is used to perform real-time compilation and spatial interpolation of the rainfall data monitored by the rain gauge station to obtain spatially interpolated monitoring rainfall data. Specifically, during the rainfall period, the collection frequency of the monitoring data of the rain gauge station is increased to the minute level, and it is updated in real time to the rainfall in the current hour. Among them, for the monitoring rainfall across the hour, the rainfall is allocated to different time periods according to the time length ratio; based on the spatial interpolation technology, the hourly monitoring rainfall data of the rain gauge station is interpolated to each grid in the region. Among them, the hourly monitoring rainfall data of the rain gauge station includes the historical compiled hourly rainfall data and the hourly rainfall data compiled and updated in real time in the current hour. Multi-source forecast rainfall data integration and correction module, used to integrate and correct multi-source short-term forecast rainfall data and multi-source medium-term forecast rainfall data respectively; The data splicing module is used to finely splice the spatially interpolated monitored rainfall data, the integrated and revised short-term forecast rainfall data, and the integrated and revised medium- and short-term forecast rainfall data in the time dimension to obtain refined rainfall data; Analysis and early warning of flash flood disasters based on refined rainfall data and rainfall warning indicators of flash flood risks.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for forecasting and warning flash floods based on refined rainfall analysis as claimed in any one of claims 1 to 6 is implemented.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the flash flood forecasting and warning method based on refined rainfall analysis as described in any one of claims 1 to 6 is implemented.

Images