A short-term heavy rainfall warning method and system integrating multi-source data

Abstract

The invention relates to the technical field of rainfall warning, and discloses a short-term heavy rainfall warning method and system integrating multi-source data. The method comprises: obtaining forecast data from a forecast data source, performing data fusion, and obtaining integrated forecast data; obtaining actual rainfall data of a regional site, and determining a target warning site; obtaining first longitude and latitude information of the target warning site and adjacent longitude and latitude information, and determining rainfall forecast data for the next 3 hours; superimposing the rainfall forecast data for the next 3 hours and the actual rainfall data of the regional site, and obtaining a rainfall data superposition coefficient; determining the township where the regional short-term heavy rainfall warning site is located according to the first longitude and longitude information and geographic information data, and generating a graded short-term heavy rainfall warning prompt according to the maximum short-term heavy rainfall data corresponding to the township, so as to realize intelligent monitoring and warning of short-term heavy rainfall, ensure the accuracy and real-time performance of the short-term heavy rainfall warning, and further realize graded rainfall rolling warning of the township.

Description

Technical Field

[0001] The present invention relates to the technical field of rainfall warning, and in particular to a short-term heavy rainfall warning method and system integrating multi-source data. Background Art

[0002] Short-term heavy precipitation refers to a weather phenomenon in which the precipitation reaches or exceeds 20 mm within 1 hour. This weather phenomenon is part of severe convective weather, which is characterized by strong suddenness, great destructive power, concentrated precipitation time and strong locality. The formation of short-term heavy precipitation is usually related to the unstable state of the atmosphere. When warm and humid air rises rapidly and cools, strong convection will form, resulting in heavy precipitation in a short period of time. In general, short-term heavy precipitation is a weather phenomenon with strong destructive power. It not only affects people's daily lives, but may also have a major impact on many fields such as agriculture and transportation. Therefore, strengthening the understanding and prevention of short-term heavy precipitation is of great significance to protecting people's lives and property.

[0003] The traditional short-term heavy rainfall warning method mainly relies on the duty personnel to look at the weather map or live radar map and other software. When a sudden weather situation occurs in a certain place, the current weather situation is sent to the relevant responsible persons and units through the information sending channel. This warning method has certain monitoring blind spots. At night or when the duty personnel are easily tired, it is impossible to monitor the short-term heavy rainfall weather in the first time, and it is impossible to spread the weather information in the first time. In addition, the method based on manual alarm cannot achieve minute-level rolling updates. Summary of the invention

[0004] In view of this, the present invention proposes a short-term heavy rainfall warning method and system that integrates multi-source data. The present invention can realize intelligent monitoring and warning of short-term heavy rainfall, ensure the accuracy and real-time nature of short-term heavy rainfall warnings, and further realize rolling warnings in towns and villages at each level.

[0005] The present invention proposes a short-term heavy rainfall warning method integrating multi-source data, comprising:

[0006] Determine a forecast data source, obtain forecast data from the forecast data source, and perform data fusion on the forecast data to obtain fused forecast data, wherein the forecast data includes minute-level forecast data and hour-level forecast data;

[0007] Determine a regional short-term heavy rainfall warning site, obtain the regional site real-time rainfall data of the regional short-term heavy rainfall warning site, and determine the target warning site according to the regional site real-time rainfall data;

[0008] Acquire the first longitude and latitude information of the target warning site, and acquire the longitude and latitude information of the neighboring area 10 km of the target warning site, and determine the future 3-hour rainfall forecast data of the target warning site based on the neighboring longitude and latitude information and the fused forecast data;

[0009] Superimposing the future 3-hour rainfall forecast data and the actual rainfall data of the regional station to obtain a rainfall data superposition coefficient;

[0010] Determine whether the rainfall data superposition coefficient meets the preset conditions. If so, determine the township where the short-term heavy rainfall warning site in the area is located based on the first longitude and latitude information and geographic information data, and generate graded short-term heavy rainfall warning prompts based on the maximum short-term heavy rainfall data corresponding to the township.

[0011] Further, when obtaining forecast data from the forecast data source and fusing the forecast data to obtain fused forecast data, it includes:

[0012] A timestamp a and a+1 are preset, where a=2h;

[0013] Based on the a timestamp, obtain from the forecast data source the grid forecast data for the next 2 hours that is updated every 10 minutes in a first time period, wherein the first time period is [0, 2h];

[0014] Based on the a+1 timestamp, obtain from the forecast data source the grid forecast data for the next 36 hours that is updated hourly in a second time period, wherein the second time period is [0, 36h];

[0015] The minute-level forecast data of the first time period and the hour-level forecast data of the second time period are spliced ​​to determine new nc format file data, and the new nc format file data is used as the fused forecast data.

[0016] Furthermore, when obtaining the regional site real-time rainfall data of the regional short-term heavy rainfall warning site and determining the target warning site according to the real-time rainfall data, it includes:

[0017] Acquire a rainfall data threshold corresponding to the regional short-term heavy rainfall warning site, and determine a target warning site according to the relationship between the actual rainfall data of the regional site and the rainfall data threshold;

[0018] When the actual rainfall data of the regional site is greater than or equal to the rainfall data threshold, the corresponding regional short-term heavy rainfall warning site is determined as the target warning site;

[0019] When the actual rainfall data of the regional site is less than the rainfall data threshold, the corresponding regional short-term heavy rainfall warning site is determined as a non-target warning site.

[0020] Further, when determining the future 3-hour rainfall forecast data of the target warning site based on the adjacent longitude and latitude information and the fused forecast data, it includes:

[0021] According to the adjacent longitude and latitude information, grid fusion forecast data corresponding to the target warning site is intercepted from the fusion forecast data, and a grid fusion forecast data set is constructed according to the grid fusion forecast data;

[0022] Calculating a data mean and a data standard deviation of the grid fusion forecast data set, and extracting grid fusion forecast data greater than the data mean and greater than the data standard deviation from the grid fusion forecast data set to generate a first discrimination number series;

[0023] Extracting grid point fusion forecast data equal to the data mean and grid point fusion forecast data equal to the data standard deviation from the grid point fusion forecast data set to generate a second discrimination number series;

[0024] Extracting grid point fusion forecast data smaller than the data mean and smaller than the data standard deviation from the grid point fusion forecast data set to generate a third discrimination number series;

[0025] Calculate the data discrimination coefficient of the grid fusion forecast data set based on the first discrimination number series, the second discrimination number series and the third discrimination number series;

[0026] Calculating a data ratio of each grid point fusion forecast data in the grid point fusion forecast data set to the data discrimination coefficient;

[0027] All data ratios are analyzed, and based on the analysis results, the future 3-hour rainfall forecast data for the target warning site is calculated.

[0028] Further, when calculating the data discrimination coefficient of the grid fusion forecast data set based on the first discrimination number series, the second discrimination number series and the third discrimination number series, it includes:

[0029] Configure a first calculation coefficient for the first distinguished number of columns, and configure a second calculation coefficient for the third distinguished number of columns;

[0030] The data discrimination coefficient of the grid fusion forecast data set is calculated according to the following formula:

[0031]

[0032] Among them, c is the data differentiation coefficient of the grid fusion forecast data set, m1 is the number of grid fusion forecast data in the first differentiation series, and e i is the fusion forecast data of the i-th grid point in the first partition sequence, k1 min is the minimum grid fusion forecast data in the first partition number series, g1 is the first calculation coefficient, g2 is the second calculation coefficient, m2 is the number of grid fusion forecast data in the third partition number series, f j is the fusion forecast data of the jth grid point in the third zone number series, k2 min is the minimum grid fusion forecast data in the third differentiation series, and h is the dynamic adjustment coefficient of the data differentiation coefficient determined according to the second differentiation series.

[0033] Furthermore, the dynamic adjustment coefficient of the data differentiation coefficient is determined according to the following steps:

[0034] Count the data quantity m3 of the second distinguished number series;

[0035] Presetting a first preset dynamic adjustment coefficient, a second preset dynamic adjustment coefficient, and a third preset dynamic adjustment coefficient;

[0036] When m3 / m1+m2<1, the first preset dynamic adjustment coefficient is used as the dynamic adjustment coefficient of the data differentiation coefficient;

[0037] When 1≤m3 / m1+m2<1.2, the second preset dynamic adjustment coefficient is used as the dynamic adjustment coefficient of the data differentiation coefficient;

[0038] When 1.2≤m3 / m1+m2, the third preset dynamic adjustment coefficient is used as the dynamic adjustment coefficient of the data differentiation coefficient.

[0039] Further, when analyzing all data ratios and calculating the future 3-hour rainfall forecast data of the target warning site based on the analysis results, it includes:

[0040] Calculate the average of all data ratios and use it as the average data ratio;

[0041] The division value of the data ratio is calculated according to the following formula;

[0042] p1=(1+q)×r;

[0043] Among them, p1 is the partition value, q is the calculation factor, the value range is [0,1], and r is the average data ratio;

[0044] Generate a lower mark for all data ratios that are less than or equal to the partition value, and calculate a lower difference between the data ratio and the partition value;

[0045] Sort all lower difference values ​​by numerical size and generate a lower difference value sequence;

[0046] Generate an upper mark for all data ratios greater than the split value, and calculate an upper difference between the split value and the data ratio;

[0047] Sort all upper difference values ​​by numerical size and generate an upper difference value sequence;

[0048] Matching the lower difference sequence with the upper difference sequence in pairs to obtain a plurality of upper and lower number sequences;

[0049] Calculating the set cluster center of the grid point fusion forecast data set according to the upper and lower series, and determining the distance from each grid point fusion forecast data in the grid point fusion forecast data set to the set cluster center;

[0050] The grid fusion forecast data corresponding to the minimum distance is selected as the future 3-hour rainfall forecast data for the target warning site.

[0051] Further, when calculating the set cluster center of the grid fusion forecast data set according to the upper and lower series, it includes:

[0052] The cluster center of the grid fusion forecast data set is calculated according to the following formula:

[0053]

[0054] Among them, s is the cluster center of the grid fusion forecast data set, n is the number of upper and lower series, t1 y is the upper difference in the yth upper and lower series, t2 y is the lower difference in the yth upper and lower series, ((t1y-t2y) 2 )min for all (t1y-t2y) 2 The minimum value in ((t1y-t2y) 2 )max for all (t1y-t2y) 2 The maximum value in u 2 for all (t1y-t2y) 2 The variance of .

[0055] Further, when judging whether the rainfall data superposition coefficient meets the preset conditions, it includes:

[0056] When the rainfall data superposition coefficient is less than the preset coefficient, it is determined that the rainfall data superposition coefficient does not meet the preset condition;

[0057] When the rainfall data superposition coefficient is greater than or equal to the preset coefficient, it is determined that the rainfall data superposition coefficient meets the preset condition.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] The present invention discloses a short-term heavy rainfall warning method and system integrating multi-source data, which comprises the following steps: obtaining forecast data from a forecast data source, performing data fusion, and obtaining integrated forecast data; obtaining actual rainfall data of a regional site of a regional short-term heavy rainfall warning site, and determining a target warning site; obtaining first longitude and latitude information of the target warning site and adjacent longitude and latitude information of a 10km adjacent area, and determining rainfall forecast data for the next 3 hours; superimposing the rainfall forecast data for the next 3 hours and the actual rainfall data of the regional site, and obtaining a rainfall data superposition coefficient; determining the township where the regional short-term heavy rainfall warning site is located according to the first longitude and longitude information and geographic information data, and generating a graded short-term heavy rainfall warning prompt according to the maximum short-term heavy rainfall data corresponding to the township, so as to realize intelligent monitoring and warning of short-term heavy rainfall, ensure the accuracy and real-time performance of the short-term heavy rainfall warning, and further realize graded rainfall rolling warning for the township.

[0060] On the other hand, the present application also provides a short-term heavy rainfall warning system integrating multi-source data, including:

[0061] The first module is used to determine a forecast data source, obtain forecast data from the forecast data source, and perform data fusion on the forecast data to obtain fused forecast data, wherein the forecast data includes minute-level forecast data and hour-level forecast data;

[0062] The second module is used to determine the regional short-term heavy rainfall warning site, obtain the regional site real-time rainfall data of the regional short-term heavy rainfall warning site, and determine the target warning site according to the regional site real-time rainfall data;

[0063] The third module is used to obtain the first longitude and latitude information of the target warning site, and obtain the longitude and latitude information of the surrounding area 10km away from the target warning site, and determine the future 3-hour rainfall forecast data of the target warning site based on the surrounding longitude and latitude information and the fused forecast data;

[0064] The fourth module is used to superimpose the future 3-hour rainfall forecast data and the actual rainfall data of the regional site to obtain a rainfall data superposition coefficient;

[0065] The fifth module is used to determine whether the rainfall data superposition coefficient meets the preset conditions. If so, the township where the short-term heavy rainfall warning station in the area is located is determined according to the first longitude and latitude information and geographic information data, and a graded short-term heavy rainfall warning prompt is generated according to the maximum short-term heavy rainfall data corresponding to the township.

[0066] It can be understood that the short-term heavy rainfall warning system and method integrating multi-source data provided above have the same beneficial effects, which will not be described in detail here. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0068] Figure 1 A schematic flow chart of a short-term heavy rainfall warning method integrating multi-source data provided in an embodiment of the present invention;

[0069] Figure 2 A schematic diagram of the structure of a short-term heavy rainfall warning system integrating multi-source data provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0070] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features described in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0071] like Figure 1 As shown, in some embodiments of the present application, this embodiment provides a short-term heavy rainfall warning method integrating multi-source data, including:

[0072] S110: determining a forecast data source, acquiring forecast data from the forecast data source, and performing data fusion on the forecast data to obtain fused forecast data, wherein the forecast data includes minute-level forecast data and hour-level forecast data;

[0073] In this embodiment, the forecast data source is the Central Meteorological Observatory.

[0074] In some embodiments of the present application, when obtaining forecast data from the forecast data source and fusing the forecast data to obtain fused forecast data, the method includes:

[0075] A timestamp a and a+1 are preset, where a=2h;

[0076] Based on the a timestamp, obtain from the forecast data source the grid forecast data for the next 2 hours that is updated every 10 minutes in a first time period, wherein the first time period is [0, 2h];

[0077] Based on the a+1 timestamp, obtain from the forecast data source the grid forecast data for the next 36 hours that is updated hourly in a second time period, wherein the second time period is [0, 36h];

[0078] The minute-level forecast data of the first time period and the hour-level forecast data of the second time period are spliced ​​to determine new nc format file data, and the new nc format file data is used as the fused forecast data.

[0079] In this embodiment, minute-level forecast data for the first time period is obtained from the forecast data source based on timestamp a, that is, minute-level forecast data [0, 2h] is obtained from the forecast data source for the first two hours, and the data spatial resolution is 1km×1km, and the temporal resolution is 10min.

[0080] In this embodiment, hourly forecast data for the second time period is obtained from the forecast data source based on the a+1 timestamp, that is, [0, 36h] hourly forecast data is obtained from the forecast data source in the 3rd hour, and the data spatial resolution is 1km×1km, and the temporal resolution is 1h.

[0081] In this embodiment, the 0-2h minute level updates the forecast data for the next 2 hours, and the 0-36h hour level updates the forecast data for the next 36 hours.

[0082] In this embodiment, the update method of the minute-level forecast data of the central station (each data is a 2-dimensional data including longitude and latitude) is as follows: for example, at the current time of 14:30 on September 22, 2024, rainfall data for 14:30-15:30 and 15:30-16:30 can be forecast. 10 minutes later, the current time is 14:40 on September 22, 2024, and the rainfall data for 14:40-15:40 and 15:40-16:40 are forecast. The update method of the hour-level forecast data of the central station (each data is a 2-dimensional data including longitude and latitude) is as follows: for example, at the current time of 14:00 on September 22, 2024, rainfall data for the next 36 hours can be forecast. 1 hour later, the current time is 15:00 on September 22, 2024, and the rainfall data for the next 36 hours after 15:00 is at this time.

[0083] In this embodiment, the nc file, full name Network Common Data Format, is a file format for storing scientific data. Before splicing the minute-level forecast data of the first time period and the hour-level forecast data of the second time period to determine the new nc format file data, first, ensure that the geographic coordinates (latitude and longitude) of the minute-level forecast data and the hour-level forecast data are consistent, and secondly, ensure that the data spatial resolution of the minute-level forecast data and the hour-level forecast data is consistent. Finally, use Python's netCDF4 library to create a new NetCDF file, create a separate variable for each hour's forecast data, and finally obtain a new nc format file.

[0084] The beneficial effect of the above technical solution is: the present invention splices the minute-level forecast data of the first time period and the hour-level forecast data of the second time period, determines the new NC format file data, and uses the new NC format file data as the fused forecast data, thereby ensuring the accurate fusion of the data and laying the foundation for the township-level short-term heavy rainfall warning.

[0085] S120: Determine a regional short-term heavy rainfall warning site, obtain the regional site real-time rainfall data of the regional short-term heavy rainfall warning site, and determine the target warning site according to the regional site real-time rainfall data;

[0086] In some embodiments of the present application, when obtaining the regional site real-time rainfall data of the regional short-term heavy rainfall warning site and determining the target warning site according to the real-time rainfall data, it includes:

[0087] Acquire a rainfall data threshold corresponding to the regional short-term heavy rainfall warning site, and determine a target warning site according to the relationship between the actual rainfall data of the regional site and the rainfall data threshold;

[0088] When the actual rainfall data of the regional site is greater than or equal to the rainfall data threshold, the corresponding regional short-term heavy rainfall warning site is determined as the target warning site;

[0089] When the actual rainfall data of the regional site is less than the rainfall data threshold, the corresponding regional short-term heavy rainfall warning site is determined as a non-target warning site.

[0090] In this embodiment, the regional short-term heavy rainfall warning site refers to a short-term heavy rainfall warning site in the local area, which can be obtained from the meteorological service website of the local area.

[0091] In this example, the current month is obtained. If the current month is May to September:

[0092] 1. The alarm method is WeChat alarm. The real-time alarm product thresholds are: rain_1h: 70, 100, 120, 150; rain_3h: 100, 150, 180, 200, 250, 300; rain_6h: 100, 150, 200, 250, 300; rain_12h: 200, 250, 280, 300, 320, 350; real-time (past 3 hours) + forecast (next 3 hours) alarm product thresholds; warn_pred: 100, 150, 180, 200.

[0093] 2. If the alarm method is telephone alarm: rain_1h:80; rain_3h:150,180,200; rain_6h:180,200,250,300; rain_12h:180,200,250,300,350; actual (past 3 hours) + forecast (next 3 hours) alarm product threshold; warn_pred:150,180.

[0094] For other months:

[0095] 1. If the alarm method is WeChat alarm, the real-time alarm product threshold is: rain_1h: 50, 80, 100; rain_3h: 50, 100, 150, 180; rain_6h: 100, 150, 180; rain_12h: 100, 150, 180; real-time (past 3 hours) + forecast (next 3 hours) alarm product threshold; warn_pred: 80, 100, 150, 180.

[0096] 2. If the alarm method is telephone alarm, the threshold of real-time alarm product is: rain_1h:70; rain_3h:100,150; rain_6h:100,150; rain_12h:150,180,200; the threshold of real-time (past 3 hours) + forecast (next 3 hours) alarm product is: warn_pred:100,150,180.

[0097] The beneficial effect of the above technical solution is that the present invention can accurately determine the target alarm site according to the relationship between the actual rainfall data of the regional site and the rainfall data threshold, thereby improving work efficiency and ensuring the real-time nature of the alarm.

[0098] S130: Acquire the first longitude and latitude information of the target warning site, and acquire the longitude and latitude information of the neighboring area 10 km of the target warning site, and determine the future 3-hour rainfall forecast data of the target warning site based on the neighboring longitude and latitude information and the fused forecast data;

[0099] In some embodiments of the present application, when determining the future 3-hour rainfall forecast data of the target warning site based on the adjacent longitude and latitude information and the fused forecast data, it includes:

[0100] According to the adjacent longitude and latitude information, grid fusion forecast data corresponding to the target warning site is intercepted from the fusion forecast data, and a grid fusion forecast data set is constructed according to the grid fusion forecast data;

[0101] Calculating a data mean and a data standard deviation of the grid fusion forecast data set, and extracting grid fusion forecast data greater than the data mean and greater than the data standard deviation from the grid fusion forecast data set to generate a first discrimination number series;

[0102] Extracting grid point fusion forecast data equal to the data mean and grid point fusion forecast data equal to the data standard deviation from the grid point fusion forecast data set to generate a second discrimination number series;

[0103] Extracting grid point fusion forecast data smaller than the data mean and smaller than the data standard deviation from the grid point fusion forecast data set to generate a third discrimination number series;

[0104] Calculate the data discrimination coefficient of the grid fusion forecast data set based on the first discrimination number series, the second discrimination number series and the third discrimination number series;

[0105] Calculating a data ratio of each grid point fusion forecast data in the grid point fusion forecast data set to the data discrimination coefficient;

[0106] All data ratios are analyzed, and based on the analysis results, the future 3-hour rainfall forecast data for the target warning site is calculated.

[0107] In this embodiment, when the grid point fusion forecast data corresponding to the target warning site is intercepted from the fusion forecast data according to the neighboring latitude and longitude information, the target warning site may include more than 270 grid point fusion forecast data, so the intercepted data is further processed.

[0108] The beneficial effect of the above technical solution is: the present invention calculates the data differentiation coefficient of the grid fusion forecast data set based on the first differentiation series, the second differentiation series and the third differentiation series, which can ensure the calculation accuracy of the data differentiation coefficient and further provide reliable data support for the determination of the future 3h rainfall forecast data of the target warning site.

[0109] In some embodiments of the present application, when calculating the data discrimination coefficient of the grid fusion forecast data set based on the first discrimination number series, the second discrimination number series, and the third discrimination number series, it includes:

[0110] Configure a first calculation coefficient for the first distinguished number of columns, and configure a second calculation coefficient for the third distinguished number of columns;

[0111] The data discrimination coefficient of the grid fusion forecast data set is calculated according to the following formula:

[0112]

[0113] Among them, c is the data differentiation coefficient of the grid fusion forecast data set, m1 is the number of grid fusion forecast data in the first differentiation series, and e i is the fusion forecast data of the i-th grid point in the first partition sequence, k1 min is the minimum grid fusion forecast data in the first partition number series, g1 is the first calculation coefficient, g2 is the second calculation coefficient, m2 is the number of grid fusion forecast data in the third partition number series, f j is the fusion forecast data of the jth grid point in the third zone number series, k2 min is the minimum grid fusion forecast data in the third differentiation series, and h is the dynamic adjustment coefficient of the data differentiation coefficient determined according to the second differentiation series.

[0114] In this embodiment, the first calculation coefficient is 0.6, and the second calculation coefficient is 0.4.

[0115] In some embodiments of the present application, the dynamic adjustment coefficient of the data differentiation coefficient is determined according to the following steps:

[0116] Count the data quantity m3 of the second distinguished number series;

[0117] Presetting a first preset dynamic adjustment coefficient, a second preset dynamic adjustment coefficient, and a third preset dynamic adjustment coefficient;

[0118] When m3 / m1+m2<1, the first preset dynamic adjustment coefficient is used as the dynamic adjustment coefficient of the data differentiation coefficient;

[0119] When 1≤m3 / m1+m2<1.2, the second preset dynamic adjustment coefficient is used as the dynamic adjustment coefficient of the data differentiation coefficient;

[0120] When 1.2≤m3 / m1+m2, the third preset dynamic adjustment coefficient is used as the dynamic adjustment coefficient of the data differentiation coefficient.

[0121] In this embodiment, the first preset dynamic adjustment coefficient is preferably 0.95, the second preset dynamic adjustment coefficient is preferably 1.15, and the third preset dynamic adjustment coefficient is preferably 1.25.

[0122] The beneficial effect of the above technical solution is that the present invention realizes the dynamic adjustment of the data differentiation coefficient, further ensures the accuracy of the data differentiation coefficient, and avoids errors.

[0123] In some embodiments of the present application, when analyzing all data ratios and calculating the future 3-hour rainfall forecast data of the target warning site based on the analysis results, it includes:

[0124] Calculate the average of all data ratios and use it as the average data ratio;

[0125] The division value of the data ratio is calculated according to the following formula;

[0126] p1=(1+q)×r;

[0127] Among them, p1 is the partition value, q is the calculation factor, the value range is [0,1], and r is the average data ratio;

[0128] Generate a lower mark for all data ratios that are less than or equal to the partition value, and calculate a lower difference between the data ratio and the partition value;

[0129] Sort all lower difference values ​​by numerical size and generate a lower difference value sequence;

[0130] Generate an upper mark for all data ratios greater than the split value, and calculate an upper difference between the split value and the data ratio;

[0131] Sort all upper difference values ​​by numerical size and generate an upper difference value sequence;

[0132] Matching the lower difference sequence with the upper difference sequence in pairs to obtain a plurality of upper and lower number sequences;

[0133] Calculating the set cluster center of the grid point fusion forecast data set according to the upper and lower series, and determining the distance from each grid point fusion forecast data in the grid point fusion forecast data set to the set cluster center;

[0134] The grid fusion forecast data corresponding to the minimum distance is selected as the future 3-hour rainfall forecast data for the target warning site.

[0135] In this embodiment, when the lower difference sequence and the upper difference sequence are matched in pairs, the maximum upper difference in the upper difference sequence is selected, the maximum lower difference in the lower difference sequence is selected, and then the two values ​​are matched, the adjacent value of the maximum upper difference in the upper difference sequence is selected, the adjacent value of the maximum lower difference in the lower difference sequence is selected, and then the two adjacent values ​​are matched, and the rest are matched in pairs in sequence according to this rule.

[0136] In this embodiment, the method for determining the distance from each grid point fusion forecast data to the set cluster center is complicated and mature, and will not be introduced in detail here to save space.

[0137] The beneficial effect of the above technical solution is that the present invention selects the station fusion forecast number corresponding to the minimum distance as the future 3-hour rainfall forecast data of the target alarm station, thereby ensuring the intelligent determination of the future 3-hour rainfall forecast data and avoiding the errors and untimeliness of manual participation.

[0138] In some embodiments of the present application, when calculating the set cluster center of the grid fusion forecast data set according to the upper and lower series, it includes:

[0139] The cluster center of the grid fusion forecast data set is calculated according to the following formula:

[0140]

[0141] Among them, s is the cluster center of the grid fusion forecast data set, n is the number of upper and lower series, t1 y is the upper difference in the yth upper and lower series, t2 y is the lower difference in the yth upper and lower series, ((t1y-t2y) 2 )min for all (t1y-t2y) 2 The minimum value in ((t1y-t2y) 2 )max for all (t1y-t2y) 2 The maximum value in u 2 for all (t1y-t2y) 2 The variance of .

[0142] S140: superimposing the rainfall forecast data for the next 3 hours and the actual rainfall data of the regional station to obtain a rainfall data superposition coefficient;

[0143] In this embodiment, the rainfall forecast data for the next 3 hours and the actual rainfall data of the regional stations are superimposed, that is, the sum of the two is calculated.

[0144] S150: Determine whether the rainfall data superposition coefficient meets the preset conditions. If so, determine the township where the short-term heavy rainfall warning site in the area is located based on the first longitude and latitude information and geographic information data, and generate graded short-term heavy rainfall warning prompts based on the maximum short-term heavy rainfall data corresponding to the township.

[0145] In some embodiments of the present application, when judging whether the rainfall data superposition coefficient meets the preset conditions, it includes:

[0146] When the rainfall data superposition coefficient is less than the preset coefficient, it is determined that the rainfall data superposition coefficient does not meet the preset condition;

[0147] When the rainfall data superposition coefficient is greater than or equal to the preset coefficient, it is determined that the rainfall data superposition coefficient meets the preset condition.

[0148] In this embodiment, Liuchuan Town, Jianhe County, Qiandongnan Prefecture is taken as an example to show the graded short-term heavy rainfall warning prompts:

[0149] [Short-term heavy rainfall forecast warning]: 2024-09-16 08:49, the maximum rainfall in Liuchuan Town, Jianhe County, Qiandongnan Prefecture in the past 3 hours was 176.3mm. In the past 3 hours, 2 monitoring stations within 10 kilometers have had rainfall exceeding 30mm, and 1 observation station has had rainfall exceeding 50mm. It is expected that the rainfall will continue in the next 3 hours, and the maximum rainfall in Jianhe County may reach more than 200.0mm. Please pay attention to prevention at the Qiandongnan Prefecture Meteorological Observatory and Jianhe County Meteorological Observatory!

[0150] [Heavy Rainfall Warning]: Liuchuan Town, Jianhe County, Qiandongnan Prefecture, 2024-09-16 08:43, precipitation in the past hour was 70.2mm. Please take precautions at the Qiandongnan Prefecture Meteorological Observatory and Jianhe County Meteorological Observatory!

[0151] [Heavy Rainfall Warning]: Liuchuan Town, Jianhe County, Qiandongnan Prefecture, 2024-09-16 08:44, precipitation in the past 3 hours was 184.6mm. Please take precautions at the Qiandongnan Prefecture Meteorological Observatory and Jianhe County Meteorological Observatory!

[0152] [Heavy Rainfall Warning]: Liuchuan Town, Jianhe County, Qiandongnan Prefecture, 2024-09-16 08:45, precipitation in the past 6 hours was 203.6mm. Please take precautions at the Qiandongnan Prefecture Meteorological Observatory and Jianhe County Meteorological Observatory!

[0153] [Heavy Rainfall Warning]: Liuchuan Town, Jianhe County, Qiandongnan Prefecture, 2024-09-16 08:42, precipitation in the past 12 hours was 227.3mm. Please take precautions at the Qiandongnan Prefecture Meteorological Observatory and Jianhe County Meteorological Observatory!

[0154] In some embodiments of the present application, an example of September 11, 2023 was selected for analysis. This process was mainly affected by the coordination of high-altitude troughs and mid- and low-level shear. Heavy precipitation occurred in the northern and western parts of Guizhou Province from 08:00 on September 11 to 08:00 on September 12. The heavy precipitation period was mainly concentrated from 20:00 on the 7th to 06:00 on the 8th, which was the period with the heaviest precipitation. The precipitation areas with large values ​​were mainly concentrated in Liupanshui City, Bijie City, and the western and northern parts of Zunyi City. The accuracy of the patented method in this process was 0.9438, the hit rate was 1, the missed rate was 0, and the lead time was 97 minutes compared with the actual situation; while the accuracy of the manual release was 0.921, the hit rate was 0.941, and the warning lead time was 59 minutes.

[0155] Around 17:00 on the 11th, convection was generated and developed in the northern part of Bijie City and the northwest part of Zunyi City. From 17:00 to 18:00, two stations in Chishui City experienced short-term heavy rainfall, with hourly rainfall intensities of 59.6 mm at Lianghekou Station and 54.1 mm at Tiantai Station, respectively. The city reached the yellow warning standard for heavy rain. From the perspective of this patent method, at 17:10, it was predicted that 12 grid points in Chishui City would reach the yellow warning standard for heavy rain, and a yellow warning for heavy rain was issued to Chishui City for the first time. From the perspective of encrypted real-time precipitation data, the hourly cumulative precipitation at Lianghekou Station in Chishui City at 18:00 was greater than 50 mm, and the 6-hour precipitation was 59.6 mm, reaching the yellow warning standard. A yellow warning signal for heavy rain should be issued. From the perspective of city and county service conditions, the Chishui Meteorological Bureau issued a yellow warning for heavy rain at 16:55. The objective product of the heavy rain warning was 50 minutes ahead of the actual situation and 15 minutes ahead of the forecaster's warning.

[0156] From 20:00 to 21:00 on the 11th, the 6-hour precipitation at Duipo Station in Qixingguan District, Bijie City was 59.7mm, and the city reached the yellow warning standard for heavy rain. From the perspective of this patent method, at 19:30, it is predicted that 5 grid points in Qixingguan District will reach the yellow warning standard for heavy rain, and the first yellow warning reminder for heavy rain was issued for Qixingguan, Bijie City. From the encrypted real-time precipitation data, the 6-hour cumulative precipitation at Duipo Station in Qixingguan District at 21:00 was greater than 50mm, reaching the yellow warning standard, and a yellow warning signal for heavy rain should be issued. From the perspective of city and county service conditions, the Qixingguan District Meteorological Bureau issued a yellow warning for heavy rain at 19:33. The objective product of the heavy rain warning was 90 minutes ahead of the actual situation and 3 minutes ahead of the forecaster's warning.

[0157] At 20:00, high pressure in the western region moved eastward. Affected by the low-level shear system, scattered convection developed in the northwestern part of the province. Due to the strong airflow on the periphery of the subtropical high, the convection was stable and less active, and the accumulated rainfall was large. From 23:00 on the 11th to 00:00 on the 12th, the 6-hour precipitation in Kaiga, Ganhe and Chashan in Weining was greater than 50 mm, and the hourly rainfall intensities were 48.4 mm, 52.3 mm and 23.4 mm respectively. The county reached the heavy rain warning standard. At 01:20 on the 12th, the 3-hour precipitation at Yuanping Station in Weining County reached 103.5 mm, reaching the heavy rain orange warning standard. The heavy rain yellow warning signal should be upgraded to heavy rain orange warning signal. According to the patent method, at 21:30, it is predicted that the precipitation at 4 grid points in the southeast and north of Weining County will reach the yellow rainstorm warning standard, and a yellow rainstorm warning should be issued. At 00:30 on the 12th, one grid point in Weining County is expected to reach the orange rainstorm warning signal, and the warning signal should be upgraded. From the actual observation data, at 00:00 on the 12th, three stations in Weining County reached the yellow rainstorm warning standard, and at 01:20, one station reached the orange rainstorm warning standard. From the perspective of city and county service conditions, the Weining County Meteorological Bureau reached the orange rainstorm warning standard at 2 A yellow rainstorm warning was issued at 1:45 am, with a warning lead time of 135 minutes in the county. The objective product of the rainstorm warning was 150 minutes ahead of the actual situation and 15 minutes ahead of the forecaster's release. Subsequently, at 00:58 am on the 12th, the Weining County Meteorological Bureau upgraded the yellow rainstorm warning signal to an orange rainstorm warning signal, with a warning lead time of 22 minutes. The objective product of the rainstorm warning was 50 minutes ahead of the actual situation and 28 minutes ahead of the forecaster's release. The objective product of the rainstorm warning has a good guiding role in city and county services.

[0158] Comparative analysis shows that this patented method has a certain lead time compared to manual warning, especially for the initial convection situation, the lead time is more obvious and has a higher hit rate.

[0159] like Figure 2 As shown, in another preferred embodiment based on the above embodiment, this embodiment provides a short-term heavy rainfall warning system integrating multi-source data, including:

[0160] The first module is used to determine a forecast data source, obtain forecast data from the forecast data source, and perform data fusion on the forecast data to obtain fused forecast data, wherein the forecast data includes minute-level forecast data and hour-level forecast data;

[0161] The second module is used to determine the regional short-term heavy rainfall warning site, obtain the regional site real-time rainfall data of the regional short-term heavy rainfall warning site, and determine the target warning site according to the regional site real-time rainfall data;

[0162] The third module is used to obtain the first longitude and latitude information of the target warning site, and obtain the longitude and latitude information of the surrounding area 10km away from the target warning site, and determine the future 3-hour rainfall forecast data of the target warning site based on the surrounding longitude and latitude information and the fused forecast data;

[0163] The fourth module is used to superimpose the future 3-hour rainfall forecast data and the actual rainfall data of the regional site to obtain a rainfall data superposition coefficient;

[0164] The fifth module is used to determine whether the rainfall data superposition coefficient meets the preset conditions. If so, the township where the short-term heavy rainfall warning station in the area is located is determined according to the first longitude and latitude information and geographic information data, and a graded short-term heavy rainfall warning prompt is generated according to the maximum short-term heavy rainfall data corresponding to the township.

[0165] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt 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 codes.

[0166] The present application is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, 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 generate 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 flowchart and / or block diagram. 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.

[0167] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0168] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0169] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A short-term heavy rainfall warning method integrating multi-source data, characterized in that: include: Determine a forecast data source, obtain forecast data from the forecast data source, and perform data fusion on the forecast data to obtain fused forecast data, wherein the forecast data includes minute-level forecast data and hour-level forecast data; Determine a regional short-term heavy rainfall warning site, obtain the regional site real-time rainfall data of the regional short-term heavy rainfall warning site, and determine the target warning site according to the regional site real-time rainfall data; Acquire the first longitude and latitude information of the target warning site, and acquire the longitude and latitude information of the neighboring area 10 km of the target warning site, and determine the future 3-hour rainfall forecast data of the target warning site based on the neighboring longitude and latitude information and the fused forecast data; Superimposing the future 3-hour rainfall forecast data and the actual rainfall data of the regional station to obtain a rainfall data superposition coefficient; Determine whether the rainfall data superposition coefficient meets the preset conditions. If so, determine the township where the regional short-term heavy rainfall warning station is located according to the first longitude and latitude information and geographic information data, and generate graded short-term heavy rainfall warning prompts according to the maximum short-term heavy rainfall data corresponding to the township; When determining the future 3-hour rainfall forecast data of the target warning site based on the adjacent latitude and longitude information and the fused forecast data, it includes: According to the adjacent longitude and latitude information, grid fusion forecast data corresponding to the target warning site is intercepted from the fusion forecast data, and a grid fusion forecast data set is constructed according to the grid fusion forecast data; Calculating a data mean and a data standard deviation of the grid fusion forecast data set, and extracting grid fusion forecast data greater than the data mean and greater than the data standard deviation from the grid fusion forecast data set to generate a first discrimination number series; Extracting grid point fusion forecast data equal to the data mean and grid point fusion forecast data equal to the data standard deviation from the grid point fusion forecast data set to generate a second discrimination number series; Extracting grid point fusion forecast data smaller than the data mean and smaller than the data standard deviation from the grid point fusion forecast data set to generate a third discrimination number series; Calculate the data discrimination coefficient of the grid fusion forecast data set based on the first discrimination number series, the second discrimination number series and the third discrimination number series; Calculating a data ratio of each grid point fusion forecast data in the grid point fusion forecast data set to the data discrimination coefficient; Analyze all data ratios, and calculate the future 3-hour rainfall forecast data of the target warning site based on the analysis results; When calculating the data discrimination coefficient of the grid fusion forecast data set based on the first discrimination number series, the second discrimination number series and the third discrimination number series, it includes: Configure a first calculation coefficient for the first distinguished number of columns, and configure a second calculation coefficient for the third distinguished number of columns; The data discrimination coefficient of the grid fusion forecast data set is calculated according to the following formula: Among them, c is the data differentiation coefficient of the grid fusion forecast data set, m1 is the number of grid fusion forecast data in the first differentiation series, and e i is the fusion forecast data of the i-th grid point in the first partition sequence, k1 min is the minimum grid fusion forecast data in the first partition number series, g1 is the first calculation coefficient, g2 is the second calculation coefficient, m2 is the number of grid fusion forecast data in the third partition number series, f j is the fusion forecast data of the jth grid point in the third zone number series, k2 min is the minimum grid point fusion forecast data in the third differentiation series, and h is the dynamic adjustment coefficient of the data differentiation coefficient determined according to the second differentiation series; Determine the dynamic adjustment factor for the data discrimination factor using the following steps: Count the data quantity m3 of the second distinguished number series; Presetting a first preset dynamic adjustment coefficient, a second preset dynamic adjustment coefficient, and a third preset dynamic adjustment coefficient; When m3 / m1+m2<1, the first preset dynamic adjustment coefficient is used as the dynamic adjustment coefficient of the data differentiation coefficient; When 1≤m3 / m1+m2<1.2, the second preset dynamic adjustment coefficient is used as the dynamic adjustment coefficient of the data differentiation coefficient; When 1.2≤m3 / m1+m2, the third preset dynamic adjustment coefficient is used as the dynamic adjustment coefficient of the data differentiation coefficient; When analyzing all data ratios and calculating the future 3-hour rainfall forecast data of the target warning site based on the analysis results, it includes: Calculate the average of all data ratios and use it as the average data ratio; The division value of the data ratio is calculated according to the following formula; p1=(1+q)×r; Among them, p1 is the partition value, q is the calculation factor, the value range is [0,1], and r is the average data ratio; Generate a lower mark for all data ratios that are less than or equal to the partition value, and calculate a lower difference between the data ratio and the partition value; Sort all lower difference values ​​by numerical size and generate a lower difference value sequence; Generate an upper mark for all data ratios greater than the split value, and calculate an upper difference between the split value and the data ratio; Sort all upper difference values ​​by numerical size and generate an upper difference value sequence; Matching the lower difference sequence with the upper difference sequence in pairs to obtain a plurality of upper and lower number sequences; Calculating the set cluster center of the grid point fusion forecast data set according to the upper and lower series, and determining the distance from each grid point fusion forecast data in the grid point fusion forecast data set to the set cluster center; The grid fusion forecast data corresponding to the minimum distance is selected as the future 3-hour rainfall forecast data for the target warning site.

2. The short-term heavy rainfall warning method integrating multi-source data according to claim 1 is characterized in that: When obtaining forecast data from the forecast data source and fusing the forecast data to obtain fused forecast data, the method includes: A timestamp a and a+1 are preset, where a=2h; Based on the a timestamp, obtain from the forecast data source the grid forecast data for the next 2 hours that is updated every 10 minutes in a first time period, wherein the first time period is [0, 2h]; Based on the a+1 timestamp, obtain from the forecast data source the grid forecast data for the next 36 hours that is updated hourly in a second time period, wherein the second time period is [0, 36h]; The minute-level forecast data of the first time period and the hour-level forecast data of the second time period are spliced ​​to determine new nc format file data, and the new nc format file data is used as the fused forecast data.

3. The short-term heavy rainfall warning method integrating multi-source data according to claim 1 is characterized in that: When obtaining the regional site real-time rainfall data of the regional short-term heavy rainfall warning site and determining the target warning site according to the real-time rainfall data, it includes: Acquire a rainfall data threshold corresponding to the regional short-term heavy rainfall warning site, and determine a target warning site according to the relationship between the actual rainfall data of the regional site and the rainfall data threshold; When the actual rainfall data of the regional site is greater than or equal to the rainfall data threshold, the corresponding regional short-term heavy rainfall warning site is determined as the target warning site; When the actual rainfall data of the regional site is less than the rainfall data threshold, the corresponding regional short-term heavy rainfall warning site is determined as a non-target warning site.

4. The short-term heavy rainfall warning method integrating multi-source data according to claim 1 is characterized in that: When calculating the set cluster center of the grid fusion forecast data set according to the upper and lower series, it includes: The cluster center of the grid fusion forecast data set is calculated according to the following formula: Among them, s is the cluster center of the grid fusion forecast data set, n is the number of upper and lower series, t1 y is the upper difference in the yth upper and lower series, t2 y is the lower difference in the yth upper and lower series, ((t1y-t2y) 2 )min for all (t1y-t2y) 2 The minimum value in ((t1y-t2y) 2 )max for all (t1y-t2y) 2 The maximum value in u 2 for all (t1y-t2y) 2 The variance of .

5. The short-term heavy rainfall warning method integrating multi-source data according to claim 1 is characterized in that: When judging whether the rainfall data superposition coefficient meets the preset conditions, it includes: When the rainfall data superposition coefficient is less than the preset coefficient, it is determined that the rainfall data superposition coefficient does not meet the preset condition; When the rainfall data superposition coefficient is greater than or equal to the preset coefficient, it is determined that the rainfall data superposition coefficient meets the preset condition.

6. A short-term heavy rainfall warning system integrating multi-source data, applied to the short-term heavy rainfall warning method integrating multi-source data as described in any one of claims 1 to 5, characterized in that: include: The first module is used to determine a forecast data source, obtain forecast data from the forecast data source, and perform data fusion on the forecast data to obtain fused forecast data, wherein the forecast data includes minute-level forecast data and hour-level forecast data; The second module is used to determine the regional short-term heavy rainfall warning site, obtain the regional site real-time rainfall data of the regional short-term heavy rainfall warning site, and determine the target warning site according to the regional site real-time rainfall data; The third module is used to obtain the first longitude and latitude information of the target warning site, and obtain the longitude and latitude information of the surrounding area 10km away from the target warning site, and determine the future 3-hour rainfall forecast data of the target warning site based on the surrounding longitude and latitude information and the fused forecast data; The fourth module is used to superimpose the future 3-hour rainfall forecast data and the actual rainfall data of the regional site to obtain a rainfall data superposition coefficient; The fifth module is used to determine whether the rainfall data superposition coefficient meets the preset conditions. If so, the township where the short-term heavy rainfall warning station in the area is located is determined according to the first longitude and latitude information and geographic information data, and a graded short-term heavy rainfall warning prompt is generated according to the maximum short-term heavy rainfall data corresponding to the township.

Images