Mountain torrent forecasting and early warning method and device based on refined rainfall analysis

Abstract

The invention discloses a torrential flood forecasting and early warning method and device based on refined rainfall analysis, and the method comprises the following steps: 1, carrying out the investigation and evaluation of torrential flood, and determining the critical rainfall of each torrential flood risk region; 2, reorganizing rainfall site monitoring rainfall data in real time and performing spatial interpolation; 3, multi-source data integration-correction processing of short-term and imminent rainfall forecast and medium and short-term rainfall forecast; step 4, carrying out real-time rolling refined splicing on the monitored rainfall of the spatial interpolation, the integrated-corrected short-term and imminent forecast rainfall and the integrated-corrected medium-short term forecast rainfall in the time dimension; step 5; and carrying out real-time rolling analysis on the mountain torrent disaster risk based on the refined rainfall data, and carrying out advanced early warning. The method has the advantages that mountain torrent disaster risks are analyzed in advance through high-time-efficiency and high-precision refined rainfall, early warning is carried out, precious time can be reserved for risk avoiding transfer, and life and property losses are reduced.

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, and are accompanied by landslides, collapses, and mud-rock flows. They are characterized by rapid disasters and strong destructiveness. Therefore, conducting research on flash flood forecasting and early warning, effectively identifying future flash flood disaster risks and conducting advance warnings, can reserve precious time for risk avoidance and evacuation, and are of great significance for reducing the loss of life and property.

[0003] There are two ways to forecast and warn flash floods: one is to analyze and warn about rainfall based on the forecast rainfall; the other is to analyze and warn about the water level based on the hydrological and hydrodynamic model to forecast floods. However, the flow generation and confluence mechanism of flash flood small watersheds is complex, and there are few applicable hydrological model algorithms; the terrain of flash flood gullies is complex and often changes greatly under the rapid scouring of high-frequency floods. It will cost a lot of money to collect and update terrain data in time to drive the calculation of hydrodynamic models; the terrain of flash flood gullies is undulating, the riverbed is narrow, and the water flow is rapid. There are no good conditions for the installation of hydrological monitoring equipment. In addition, the compilation of hydrological data (especially flow data) requires the investment of professional technicians. It is extremely difficult to have sufficient and high-quality hydrological data to support the calibration and verification analysis of hydrological and hydrodynamic models. In view of the coverage of flash flood risk areas in my country, it is extremely difficult to promote and apply flash flood forecast and warning based on hydrological and hydrodynamic models 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 forecast and warning method based on refined rainfall analysis, to perform refined processing on monitored, short-term forecast, and medium- and short-term forecast rainfall, to carry out flash flood forecast and warning analysis in combination with rainfall warning indicators, to identify future risks and provide advance warnings, to improve the warning analysis effect, and to reserve time for risk avoidance and evacuation.

[0005] In order to achieve the above-mentioned object, the first aspect of the present invention provides a flash flood forecasting and warning method based on refined rainfall analysis, comprising: 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; Real-time compilation and spatial interpolation of rainfall data monitored by rainfall stations are performed to obtain spatially interpolated monitoring rainfall data; 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; 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 are finely spliced ​​in the time dimension to obtain refined rainfall data; Analysis and warning of flash flood disasters based on refined rainfall data and rainfall warning indicators of flash flood risks.

[0006] In one implementation, data collection and processing are performed on flash flood risk areas to determine the critical rainfall of each flash flood risk area, including: S11: Investigate and collect historical flood data of the control sections in the 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 rainwater collection 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 by adopting the historical flood analysis method, and the step S13 is repeated to perform repeated calculations 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 to be transferred and the rainfall to be immediately transferred in different periods are obtained as rainfall warning indicators for flash flood risks.

[0007] In one embodiment, real-time compilation and spatial interpolation of rainfall data monitored by a rainfall station are performed to obtain spatially interpolated monitoring rainfall data, including: During the rainfall period, the frequency of collecting monitoring data at the rainfall station is increased to the minute level, and it is updated in real time to the rainfall in the hour. Among them, for the monitored rainfall across the hour, the rainfall is allocated to different time periods according to the proportion of time length; The hourly monitoring rainfall data of the rain gauge station is interpolated to each grid in the region based on the spatial interpolation technology, wherein the hourly monitoring rainfall data of the rain gauge station includes the historically compiled hourly rainfall data and the hourly rainfall data compiled and updated in real time in the current hour.

[0008] In one implementation, the multi-source short-term forecast rainfall data and the multi-source medium-term forecast rainfall data are integrated and corrected respectively, including: Real-time connection with multi-source short-term forecast rainfall data and multi-source medium- and short-term forecast rainfall data; Resample the multi-source short-term forecast rainfall data and the multi-source medium-term forecast rainfall data, calculate the indicators used to evaluate the data accuracy of the forecast rainfall products of each source, and calculate the weight coefficients of the integration of the forecast rainfall products of 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.

[0009] In one implementation, the correction processing of the integrated short-term forecast rainfall data and the integrated medium-term forecast rainfall data includes using a frequency matching method for correction processing, 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 station location; The collected sample data is used to calculate a set of rainfall thresholds The 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, and then the forecast rainfall value of the grid is 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.

[0010] In one embodiment, the spatially interpolated monitored rainfall data, the integrated and corrected short-term forecast rainfall data, and the integrated and corrected 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 medium- and short-term forecast rainfall data to the same spatial resolution; The resampled short-term forecast rainfall data are compiled into hourly rainfall data, including the short-term forecast rainfall data for the remaining period of this hour, the entire period of the next hour, and part of the period of the next hour; According to the ratio of the remaining time period of the next hour to the complete time 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 time period of the next hour; The rainfall monitoring data of historical time periods and part of the hour with the same spatial resolution, the integrated and corrected short-term forecast rainfall data of the remaining period of the hour, the next hour, and part of the next hour, and the integrated and corrected 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 72h forecast.

[0011] In one embodiment, flash flood disaster analysis and warning based on refined rainfall data and flash flood risk rainfall warning indicators include: 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 from the first preset period in the past to the second preset period in the future in the rainwater collection area of ​​each flash flood risk area 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, the flash flood disaster risk in the future period is analyzed in real time and rollingly; Use modern communication technology to send targeted information on flash flood disaster risks in the future.

[0012] 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: The data collection and processing module is used to collect and process data in 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 rainfall station to obtain spatially interpolated monitoring rainfall data; The multi-source forecast rainfall data integration and correction module is used to integrate and correct the multi-source short-term forecast rainfall data and the multi-source medium-term and short-term forecast rainfall data respectively; The data splicing module is used to finely splice the spatially interpolated monitoring rainfall data, the integrated and corrected short-term forecast rainfall data, and the integrated and corrected 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 early warning indicators of flash flood risk Based on the same inventive concept, the third aspect of the present invention provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the method for predicting and warning flash floods based on refined rainfall analysis described in the first aspect is implemented.

[0013] 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 executable on the processor, wherein 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.

[0014] Compared with the prior art, the advantages and beneficial technical effects of the present invention are as follows: The present invention provides a flash flood forecasting and early warning method based on refined rainfall analysis. Firstly, data of flash flood risk areas are collected and processed to determine the critical rainfall of each flash flood risk area, and further rainfall early warning indicators of flash flood risks are obtained. Multi-source short-term forecast rainfall data and multi-source medium-term and short-term forecast rainfall data are integrated and corrected respectively, so as to 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 and short-term forecast rainfall are finely spliced ​​in the time dimension, so as to ensure the high timeliness and high accuracy of rainfall data in all time periods. Based on the refined forecast rainfall and the rainfall early warning indicators for flash flood risks, the flash flood disaster risk in the next 24 hours is analyzed and advanced early warning is carried out, so as to improve the early warning effect, reserve precious time for the risk avoidance and transfer of flash flood disasters, and reduce the loss of life and property. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 creative work.

[0016] Figure 1 It is a flow chart of a flash flood forecasting and warning method based on refined rainfall analysis disclosed in an embodiment of the present invention; Figure 2 A comparison diagram of integrated-corrected forecast rainfall and monitored rainfall in an embodiment of the present invention; Figure 3 It is a schematic diagram of a flash flood forecast and warning device based on refined rainfall analysis in an embodiment of the present invention; Figure 4 A schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of the present invention; Figure 5 The figure is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The invention discloses a flash flood forecasting and early warning method based on refined rainfall analysis, comprising the following steps: step one: conducting flash flood investigation and evaluation, determining the critical rainfall of each flash flood risk area, and obtaining the rainfall early warning index of the flash flood risk according to 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-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 refined rainfall data and the rainfall early warning index of the flash flood risk, and advance early warning. The present invention has the advantages of analyzing flash flood disaster risks in advance and issuing early warnings with high-efficiency and high-precision refined rainfall, which can reserve precious time for risk avoidance and evacuation, and reduce the loss of life and property (the high timeliness is reflected in the refined splicing of three types of rainfall data in the time dimension; the high precision is reflected in the integration of multi-source data-correction processing to improve rainfall accuracy).

[0018] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 creative work are within the scope of protection of the present invention.

[0019] Embodiment 1 This embodiment discloses a flash flood forecasting and warning method based on refined rainfall analysis, see Figure 1 ,include: 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; S2: Real-time compilation and spatial interpolation of rainfall data monitored by rainfall stations to obtain spatially interpolated monitoring rainfall data; S3: Integrate and correct multi-source short-term forecast rainfall data and multi-source medium- and short-term forecast rainfall data respectively; S4: The spatially interpolated monitoring rainfall data, the integrated and corrected short-term forecast rainfall data, and the integrated and corrected medium- and short-term forecast rainfall data are finely spliced ​​in the time dimension to obtain the refined rainfall data; S5: Analysis and warning of flash flood disasters based on refined rainfall data and rainfall warning indicators of flash flood risks.

[0020] Specifically, through analyzing the 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 precise rainfall data. At present, my country has accelerated the construction of a rainfall monitoring and reporting system, and the fine-tuning of rainfall data on this basis can improve the warning effect.

[0021] The spatial interpolation method in S2 can adopt the inverse distance weighted method, Kriging method, etc. Short-term forecast rainfall data generally refers to rainfall data with a forecast period of 2 hours, and medium-term and short-term forecast rainfall data generally refers to rainfall data with a forecast period of 3 days or more.

[0022] In one implementation, data collection and processing are performed on flash flood risk areas to determine the critical rainfall of each flash flood risk area, including: S11: Investigate and collect historical flood data of the control sections in the 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 rainwater collection 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; S14: Based on the design flood process of the control section, the critical water level is determined by adopting the historical flood analysis method, and the step S13 is repeated to perform repeated calculations 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 to be transferred and the rainfall to be immediately transferred in different periods are obtained as rainfall warning indicators for flash flood risks.

[0023] Specifically, S12 performs a series of operations on the digital elevation model, such as digital river channel correction, depression filling, flow direction generation, and grid rainwater collection area calculation, to extract the rainwater collection area of ​​the control section. The flood process calculated in S13 includes water level and flow process. The critical rainfall in S14 includes 1h critical rainfall, 3h critical rainfall, 6h critical rainfall, etc.

[0024] In one embodiment, real-time compilation and spatial interpolation of rainfall data monitored by a rainfall station are performed to obtain spatially interpolated monitoring rainfall data, including: During the rainfall period, the frequency of collecting monitoring data at the rainfall station is increased to the minute level, and it is updated in real time to the rainfall in the hour. Among them, for the monitored rainfall across the hour, the rainfall is allocated to different time periods according to the proportion of time length; The hourly monitoring rainfall data of the rain gauge station is interpolated to each grid in the region based on the spatial interpolation technology, wherein the hourly monitoring rainfall data of the rain gauge station includes the historically compiled hourly rainfall data and the hourly rainfall data compiled and updated in real time in the current hour.

[0025] 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 in the period 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.

[0026] 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.

[0027] 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.

[0028] In one implementation, the multi-source short-term forecast rainfall data and the multi-source medium-term forecast rainfall data are integrated and corrected respectively, including: Real-time connection with multi-source short-term forecast rainfall data and multi-source medium- and short-term forecast rainfall data; Resample the multi-source short-term forecast rainfall data and the multi-source medium-term forecast rainfall data, calculate the indicators used to evaluate the data accuracy of the forecast rainfall products of each source, and calculate the weight coefficients of the integration of the forecast rainfall products of 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.

[0029] Specifically, multi-source short-term forecast rainfall data and multi-source medium- and short-term forecast rainfall data are multi-source short-term forecast rainfall products and multi-source medium- and short-term forecast rainfall products. Generally, for multi-source short-term forecast rainfall data (forecast period is 2h), the temporal resolution, spatial resolution and update frequency of each source data are relatively consistent, which are 5min, 1km×1km and updated every 5min respectively.

[0030] For rainfall data of multi-source short-term and medium-term forecasts (forecast period of 3 days and above), such as ECMWF produced by the European Centre for Medium-Range Weather Forecasts (with a time resolution of 1 hour and a spatial resolution of 9 km × 9 km for the first 3 days, updated every 12 hours), GFS produced by the National Oceanic and Atmospheric Administration of the United States (with a time resolution of 1 hour and a spatial resolution of 25 km × 25 km for the first 3 days, updated every 6 hours), CMA produced by the China Meteorological Administration, etc. (forecast period of 3 days, time resolution of 1 hour, spatial resolution of 2 km × 2 km, updated every 12 hours).

[0031] Integrate the multi-source forecast rainfall data for short-term nowcasting and medium-term forecasts respectively to improve the stability of rainfall forecasts and reduce the impact of instability of a certain forecast rainfall data in specific scenarios. The specific steps for integrating the multi-source forecast rainfall data for short-term nowcasting and medium-term forecasts are as follows: (1) Resample the forecast rainfall data (multi-source forecast rainfall data for short-term nowcasting and medium-term forecasts) to make the spatial resolutions of each source data consistent; (2) Sample 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 size is 1 day. The sample collection content includes the monitored rainfall at the rain gauge location and the multi-source forecast rainfall.

[0032] (3) Use the sampled samples to evaluate the data accuracy of each source's forecast rainfall product 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, probability of detection POD, false alarm rate FAR, etc.), which is used as the basis for integration. Among them, the RMSE calculation formula is:

[0033] In the formula, is the forecast rainfall, is the monitored rainfall at the rain gauge; is the number of samples.

[0034] (4) Calculate the weight coefficients for integrating each source's forecast rainfall product, and integrate the multi-source forecast rainfall data set accessed in real time into a new forecast rainfall data. If integrating the multi-source forecast rainfall data based on RMSE, the weight coefficient for integrating the kth forecast rainfall product is calculated using the following formula:

[0035] The rainfall amount P of the integrated forecast rainfall product is:

[0036] In the formula, is the rainfall data of each source's forecast rainfall product.

[0037] In one implementation, the correction process for the integrated short-term rainfall forecast data and the integrated medium- and short-term rainfall forecast data includes using the frequency matching method for correction, specifically as follows: Sampling is performed based on a time-sliding window, and the sample collection content includes the monitored rainfall at the rain gauge location and the integrated forecast rainfall; For the collected sample data, a set of rainfall thresholds is statistically calculated The corresponding cumulative frequency of the integrated forecast rainfall And the cumulative frequency of the monitored rainfall , where is the number of sample rainfall data, is the Nth rainfall threshold, is The corresponding cumulative frequency of rainfall, is The corresponding cumulative frequency of the monitored rainfall; Based on linear interpolation or quadratic sample interpolation method, respectively, through the coordinates , coordinates Construct an integrated forecast rainfall cumulative frequency curve and a monitored rainfall cumulative frequency curve; For the forecast rainfall value of any grid in the integrated short-term rainfall forecast data or the integrated medium- and short-term rainfall forecast data, the cumulative frequency is obtained through the forecast rainfall cumulative frequency curve, and the monitored rainfall value corresponding to this cumulative frequency is obtained according to the monitored rainfall cumulative frequency curve. Then, the forecast rainfall value of the grid is corrected to the obtained monitored rainfall value, and finally, the integrated and corrected short-term rainfall forecast data or the integrated and corrected medium- and short-term rainfall forecast data is obtained.

[0038] Specifically, real-time rolling correction processing is performed on the integrated short-term and medium- and short-term rainfall forecasts respectively to reduce the error of the forecast rainfall and improve the forecast accuracy. The real-time rolling correction processing can adopt the frequency matching method. This method uses the deterministic forecast during the training period and the corresponding observed data (i.e., the rainfall monitored by the rain gauge station) to establish a one-to-one mapping relationship between the original forecast and the forecast correction value. The purpose is to eliminate the frequency deviation at different precipitation levels and make the total area of the forecast and observed rain areas consistent. When sampling based on the time-sliding window, the window range is one month before and after the same period of previous years, and the window step size is 1 hour.

[0039] In other implementations, other methods can be used for correction processing, such as geostatistical analysis method, LSTM method, etc.

[0040] In the specific implementation process, the upper and lower limits of the correction are set to 50 and 0.1 mm / h, respectively, to prevent excessive correction and light rain from disappearing due to extreme precipitation extrapolation. When the original forecast is greater than the upper limit of the correction, the original value is retained without correction; when the revised forecast is less than the lower limit of the correction, it is set to zero rainfall.

[0041] In one embodiment, the spatially interpolated monitored rainfall data, the integrated and corrected short-term forecast rainfall data, and the integrated and corrected 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 medium- and short-term forecast rainfall data to the same spatial resolution; The resampled short-term forecast rainfall data are compiled into hourly rainfall data, including the short-term forecast rainfall data for the remaining period of this hour, the entire period of the next hour, and part of the period of the next hour; According to the ratio of the remaining time period of the next hour to the complete time 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 time period of the next hour; The rainfall monitoring data of historical time periods and part of the hour with the same spatial resolution, the integrated and corrected short-term forecast rainfall data of the remaining period of the hour, the next hour, and part of the next hour, and the integrated and corrected 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 72h forecast.

[0042] Specifically, the spatial resolution of the resampling can be set to 1km×1km.

[0043] 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.

[0044] In one embodiment, flash flood disaster analysis and warning based on refined rainfall data and flash flood risk rainfall warning indicators include: 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 from the first preset period in the past to the second preset period in the future in the rainwater collection area of ​​each flash flood risk area 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, the flash flood disaster risk in the future period is analyzed in real time and rollingly; Use modern communication technologies to targetedly send flash flood disaster risk information for a period of time in the future.

[0045] Specifically, considering that the time periods for flash flood rainfall early warning analysis are generally 1h, 3h, and 6h, the first preset time period is set to 5h, that is, intercept the time period rainfall data of the past 5h (for flash flood rainfall early warning analysis) from the refined rainfall data; considering that the forecast rainfall error for the next 24h is relatively small, the second preset time period is set to 24h, that is, intercept the time period rainfall data of the next 24h from the refined rainfall data. In this embodiment, the surface time period rainfall from the past first preset time period to the future second preset time period represents the surface time period rainfall from the past 5h to the future 24h. Based on the rasterized data of refined rainfall, calculate the surface time period rainfall of the catchment area of each flash flood risk area for "the past 5h - the future 24h". If it is impossible to collect flash flood gully vector data or the vector data of the catchment area of the flash flood risk area, it is necessary to obtain the time period rainfall data of the corresponding raster of the flash flood risk area (or its associated rainfall station). Based on the "past 5h - the future 24h" refined time period rainfall data (catchment area surface data or flash flood risk area grid point data), combined with the ready-to-transfer rainfall and immediate-transfer rainfall for different time periods (1h, 3h, 6h), analyze the flash flood disaster risk for the next 24h in real-time and rolling manner, including the risk level and the time of risk occurrence. At the same moment, based on the flash flood disaster analysis results for different time periods, there may be differences. In this case, it is judged as the maximum possible risk.

[0046] Use modern communication technologies (such as SMS service, LBS technology (Location Based Services), virtual electronic fence, remote shouting, etc.) to targetedly send the flash flood disaster risk information for the next 24h to the flash flood responsible persons, so as to reserve transfer time for flash flood avoidance.

[0047] Now, take the flash flood early warning analysis of a certain district / county in Chongqing as an example to elaborate on the present invention in detail, which also has guiding significance for the flash flood monitoring and early warning of the present invention applied to other regions.

[0048] Referring to the attached drawings, it can be seen that: as Figure 1 shown, in this embodiment, the flash flood forecast and early warning method based on refined rainfall analysis includes the following steps: (1) Conduct flash flood dispatch evaluation on more than 100 flash flood risk areas. Through calculations such as catchment area extraction, design rainstorm, design flood (runoff yield and concentration), water level - discharge relationship, etc., obtain the design flood process of the control section, and conduct repeated trial calculations based on historical flood analysis to analyze and determine the ready-to-transfer rainfall and immediate-transfer rainfall for different time periods (1h, 3h, 6h) as the rainfall early warning indicators for flash flood risk. (2) During the rainfall period, the frequency of collecting monitoring data at the rain gauge station is increased to the minute level, and it is updated in real time to the rainfall in the hour period; for the monitored rainfall across the hour, the rainfall can be allocated to different periods according to the time length ratio; based on the inverse distance weighted method, the hourly monitored rainfall at the rain gauge station (including the historical hourly rainfall that has been compiled and the hourly rainfall that is compiled and updated in real time in the hour) is interpolated to each grid in the region, with a grid resolution of 1 km × 1 km; (3) In real time, the Caiyun short-term forecast rainfall data (with a time resolution of 5 min, a spatial resolution of 1 km × 1 km, and updated every 5 min) are connected with multi-source short-term and medium-term forecast rainfall data (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 with a range of one month before and after the same period of previous years and a step length of 1 hour, the monitored rainfall, Caiyun short-term forecast rainfall, CMA short-term and medium-term forecast rainfall, and ECMWF short-term and medium-term forecast rainfall sample data at the rain gauge location are obtained; based on the RMSE evaluation results of the forecast rainfall products of 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 short-term and medium-term forecast rainfall are corrected in real time to obtain the integrated-corrected short-term forecast rainfall and the integrated-corrected short-term forecast rainfall (see Figure 2 ); (4) Resample the integrated and revised short-term and medium-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 this hour (after the current moment), the entire period of the next hour, and part of the period of the next hour; allocate the rainfall for the next hour of the medium-term and short-term forecast according to the ratio of the remaining period of the next hour to the complete period, and obtain the short-term and medium-term forecast rainfall data for the remaining period of the next hour; and finely splice the rainfall monitoring data of the historical period and part of the current hour (before the current moment), the integrated and revised short-term forecast rainfall data for the remaining period of this hour (after the current moment), the next hour, and part of the next hour, and the integrated and revised short-term forecast rainfall data for the remaining period of the next hour and every hour thereafter in the time dimension to obtain the refined period rainfall data covering the “historical long-series monitoring - forecast for the next 72 hours”, so as to ensure that the rainfall data has extremely high timeliness and numerical accuracy.

[0049] (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 prepared 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 in a targeted manner, so as to reserve transfer time for flash flood risk avoidance.

[0050] 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 make advance warning, which can reserve precious time for risk avoidance and transfer, reduce the loss of life and property, and has significant social and economic benefits.

[0051] Embodiment 2 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: The data collection and processing module 101 is used to collect and process data of flash flood risk areas, determine the critical rainfall of each flash flood risk area, and obtain the rainfall warning index of flash flood risk according to the critical rainfall of the flash flood risk area; The monitoring rainfall compilation and interpolation module 102 is used to perform real-time compilation and spatial interpolation of the monitoring rainfall data of the rainfall station to obtain spatially interpolated monitoring rainfall data; The multi-source forecast rainfall data integration and correction module 103 is used to integrate and correct the multi-source short-term forecast rainfall data and the multi-source medium-term forecast rainfall data respectively; The data splicing module 104 is used to perform fine splicing of the spatially interpolated monitoring rainfall data, the integrated and corrected short-term forecast rainfall data, and the integrated and corrected medium-term forecast rainfall data in the time dimension to obtain refined rainfall data; 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, and evaluate its authenticity and credibility by analyzing the feature distribution visibility, detail rationality, and distortion of the enhanced image. At the same time, by embedding digital watermarks and performing image content monitoring, image data with anti-counterfeiting characteristics is generated.

[0052] Since the device introduced in the second embodiment of the present invention is the device adopted for the method of flash flood forecasting and early warning based on refined rainfall analysis in the first embodiment of the present invention, based on the method introduced in the first embodiment of the present invention, those skilled in the art can understand the specific structure and variations of the device, so it will not be elaborated here. Any device adopted by the method in the first embodiment of the present invention falls within the scope of protection of the present invention.

[0053] Embodiment Three Based on the same inventive concept, please refer to Figure 4 , the present invention also provides a computer-readable storage medium 300, on which a computer program 311 is stored, and when the program is executed by a processor, it implements the method described in Embodiment One.

[0054] Since the computer-readable storage medium introduced in the third embodiment of the present invention is the computer-readable storage medium adopted for the method of flash flood forecasting and early warning based on refined rainfall analysis in the first embodiment of the present invention, based on the method introduced in the first embodiment of the present invention, those skilled in the art can understand the specific structure and variations of the computer-readable storage medium, so it will not be elaborated here. Any computer-readable storage medium adopted by the method in the first embodiment of the present invention falls within the scope of protection of the present invention.

[0055] Embodiment Four Based on the same inventive concept, please refer to Figure 5 , the present invention also provides a computer device, including a memory 401, a processor 402, and a computer program 403 stored on the memory and executable on the processor. When the processor executes the program, it implements the method described in Embodiment One.

[0056] Since the computer device introduced in the fourth embodiment of the present invention is the computer device adopted for the method of flash flood forecasting and early warning based on refined rainfall analysis in the first embodiment of the present invention, based on the method introduced in the first embodiment of the present invention, those skilled in the art can understand the specific structure and variations of the computer device, so it will not be elaborated here. Any computer device adopted by the method in the first embodiment of the present invention falls within the scope of protection of the present invention.

[0057] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0058] The present invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each flow and / or block in the flowchart illustrations and / or block diagrams, and combinations of flows and / or blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions executed by the processor of the computer or other programmable data processing apparatus create means for implementing the functions specified in the flow Figure 1 one or more flows and / or blocks Figure 1 or means for implementing the functions specified in one or more boxes or blocks.

[0059] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made by those skilled in the art once they learn of the basic inventive concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present invention. Obviously, those skilled in the art can make various changes and variations to the embodiments of the present invention without departing from the spirit and scope of the embodiments of the present invention. Thus, if these modifications and variations of the embodiments of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A mountain flood forecasting and early warning method based on refined rainfall analysis, characterized in that, include: 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; Real-time compilation and spatial interpolation of rainfall data monitored by rainfall stations are performed to obtain spatially interpolated monitoring rainfall data; 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; 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 are finely spliced ​​in the time dimension to obtain refined rainfall data; Analysis and warning of flash flood disasters based on refined rainfall data and rainfall warning indicators of flash flood risks.

2. The flash flood forecasting and early warning method based on refined rainfall analysis according to claim 1, wherein 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 the control sections in the 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 rainwater collection 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 by adopting the historical flood analysis method, and the step S13 is repeated to perform repeated calculations 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 to be transferred and the rainfall to be immediately transferred in different periods are obtained as rainfall warning indicators for flash flood risks.

3. The flash flood forecasting and early warning method based on refined rainfall analysis according to claim 1, wherein The rainfall data monitored by the rainfall station is compiled and spatially interpolated in real time to obtain spatially interpolated monitoring rainfall data, including: During the rainfall period, the frequency of collecting monitoring data at the rainfall station is increased to the minute level, and it is updated in real time to the rainfall in the hour. Among them, for the monitored rainfall across the hour, the rainfall is allocated to different time periods according to the proportion of time length; The hourly monitoring rainfall data of the rain gauge station is interpolated to each grid in the region based on the spatial interpolation technology, wherein the hourly monitoring rainfall data of the rain gauge station includes the historically compiled hourly rainfall data and the hourly rainfall data compiled and updated in real time in the current hour.

4. The flash flood forecasting and early warning method based on refined rainfall analysis according to claim 1, wherein, The multi-source short-term forecast rainfall data and the multi-source medium- and short-term forecast rainfall data are integrated and corrected respectively, including: Real-time connection with multi-source short-term forecast rainfall data and multi-source medium- and short-term forecast rainfall data; Resample the multi-source short-term forecast rainfall data and the multi-source medium-term forecast rainfall data, calculate the indicators used to evaluate the data accuracy of the forecast rainfall products of each source, and calculate the weight coefficients of the integration of the forecast rainfall products of 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.

5. The flash flood forecasting and early warning method based on refined rainfall analysis according to claim 4, wherein, Correction processing of the integrated short-term forecast rainfall data and the integrated medium-term forecast rainfall data includes correction processing by 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 station location; The collected sample data is used to statistically calculate a set of rainfall thresholds The corresponding integrated forecast rainfall cumulative frequency and the monitored rainfall cumulative frequency , where is the quantity of the sample rainfall data is the Nth rainfall threshold is the corresponding rainfall cumulative frequency is the corresponding monitored rainfall cumulative frequency; Based on the linear interpolation or quadratic sample interpolation method, respectively construct the integrated forecast rainfall cumulative frequency curve and the monitored rainfall cumulative frequency curve through coordinate , coordinate . 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, and then the forecast rainfall value of the grid is 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.

6. The flash flood forecasting and early warning method based on refined rainfall analysis according to claim 1, wherein 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 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 medium- and short-term forecast rainfall data to the same spatial resolution; The resampled short-term forecast rainfall data are compiled into hourly rainfall data, including the short-term forecast rainfall data for the remaining period of this hour, the entire period of the next hour, and part of the period of the next hour; According to the ratio of the remaining time period of the next hour to the complete time 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 time period of the next hour; The rainfall monitoring data of historical time periods and part of the hour with the same spatial resolution, the integrated and corrected short-term forecast rainfall data of the remaining period of the hour, the next hour, and part of the next hour, and the integrated and corrected 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 72h forecast.

7. The flash flood forecasting and warning method based on refined rainfall analysis according to claim 2, wherein Analysis and warning of flash flood disasters 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 from the first preset period in the past to the second preset period in the future in the rainwater collection area of ​​each flash flood risk area 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, the flash flood disaster risk in the future period is analyzed in real time and rollingly; Use modern communication technology to send targeted information on flash flood disaster risks in the future.

8. A flash flood forecasting 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 in 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 rainfall station to obtain spatially interpolated monitoring rainfall data; The multi-source forecast rainfall data integration and correction module is used to integrate and correct the multi-source short-term forecast rainfall data and the multi-source medium-term and short-term forecast rainfall data respectively; A data splicing module, which is used to perform refined splicing of the spatially interpolated monitored rainfall data, the integrated and corrected short-term forecast rainfall data, and the integrated and corrected medium- and short-term forecast rainfall data in the time dimension to obtain refined rainfall data; Based on the refined rainfall data and the rainfall warning index of the mountain flood risk, analyze and warn of mountain flood disasters.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, it implements the mountain flood forecasting and warning method based on refined rainfall analysis as described in any one of claims 1 to 7.

10. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the mountain flood forecasting and warning method based on refined rainfall analysis as described in any one of claims 1 to 7.

Images