Strong convection radar identification product quality inspection and evaluation method, device and equipment and storage medium

Abstract

The embodiment of the invention provides a severe convection radar identification product quality inspection and evaluation method, device and equipment and a storage medium, and is applied to the technical field of meteorological monitoring. The method comprises the following steps: acquiring a severe convection radar identification product; acquiring severe convection live data in a radar single station coverage range; performing time matching and space matching on the severe convection radar identification product and the severe convection live data; and performing quality evaluation on the severe convection radar identification product according to a space-time matching result. In this way, simple and efficient quantitative and qualitative quality evaluation of severe convection radar identification products can be realized.

Description

Technical Field

[0001] The present disclosure relates to the technical field of meteorological monitoring, and particularly to a method, device, equipment, and storage medium for quality inspection and evaluation of severe convective radar recognition products. Background Art

[0002] Severe convective weather refers to disastrous weather accompanied by thunderstorm phenomena, including short-term heavy precipitation, thunderstorm gales, hail, and tornadoes, etc. Severe convective weather is mainly caused by meso-scale convective systems, which release a large amount of energy in a short period of time, forming strong convective phenomena. Severe convective weather is characterized by a small range, rapid development, and fast movement speed. It is usually accompanied by strong winds, rain, hail and other phenomena, with extremely great destructive power. These phenomena often lead to local sudden disasters, such as house collapses and casualties. Timely and accurate alarm is the key to reducing disaster losses. Weather radar is an important tool for monitoring severe convective weather, which can provide minute-level forecasts and is the most comprehensive and rapid monitoring means. Therefore, it is particularly important to timely, quickly and accurately evaluate the accuracy of severe convective radar recognition products, improve the timely and accurate monitoring and alarm of severe convective weather, and effectively reduce the losses and impacts brought by severe convective weather. Summary of the Invention

[0003] The present disclosure provides a method, device, equipment, and storage medium for quality inspection and evaluation of severe convective radar recognition products.

[0004] According to a first aspect of the present disclosure, a method for quality inspection and evaluation of severe convective radar recognition products is provided. The method includes:

[0005] Obtain a severe convective radar recognition product;

[0006] Obtain the severe convective actual situation data within the coverage of a single radar station;

[0007] Perform time matching and spatial matching on the severe convective radar recognition product and the severe convective actual situation data;

[0008] Evaluate the quality of the severe convective radar recognition product according to the spatio-temporal matching result.

[0009] In the above aspect and any possible implementation manner, a further implementation manner is provided, where the severe convective radar recognition product is a user alarm information product;

[0010] The severe convective actual situation data is the location where severe convection occurs, the start time and end time, and the intensity level;

[0011] Performing time matching and spatial matching on the user alarm information product and the severe convective actual situation data includes:

[0012] Determine a time window according to the start time and / or end time;

[0013] Taking the location where the severe convection occurs as the center and a preset length as the radius, determining the meteorological live area within the time window as the live occurrence area;

[0014] Determine whether there is alarm information issued by a user alarm information product in the live occurrence area, and if so, the match is successful.

[0015] According to the above aspects and any possible implementation, an implementation is further provided.

[0016] The quality of user alarm information products is evaluated based on the time-space matching results, including:

[0017] Determine the hit rate of the user alarm information product based on the number of successful matches;

[0018] and / or, determining the missed alarm rate and false alarm rate of the user alarm information product according to the number of matching failures;

[0019] and / or, determining the advance amount of the alarm time according to the earliest time of occurrence of severe convection in real time and the first alarm time of the user alarm information product;

[0020] And / or, based on the storm cell moving direction and speed given by the user alarm information product, the moving path is deduced, and it is determined whether the corresponding severe convection occurs in the moving path area; the correctness of the moving path assessment of the user alarm information product is determined based on the judgment result.

[0021] According to the above aspects and any possible implementation, an implementation is further provided.

[0022] The user alarm information products are divided into four categories: hail, thunderstorm gale, short-term heavy rainfall and tornado.

[0023] According to the above aspects and any possible implementation manner, an implementation manner is further provided, wherein the method further includes:

[0024] The severe convection live data is preprocessed.

[0025] According to a second aspect of the present disclosure, a device for inspecting and evaluating the quality of products identified by strong convection radar is provided. The device comprises:

[0026] Data acquisition module, used to obtain severe convection radar identification products;

[0027] The data acquisition module is also used to obtain real-time data of severe convection within the coverage area of ​​a single radar station;

[0028] A matching module for performing time matching and spatial matching on the severe convective radar identification product and the severe convective actual data;

[0029] An evaluation module for evaluating the quality of the severe convective radar identification product according to the spatio-temporal matching result.

[0030] According to a third aspect of the present disclosure, an electronic device is provided. The electronic device includes: a memory and a processor, a computer program is stored on the memory, and when the processor executes the program, the method as described above is implemented.

[0031] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the method as described in the first aspect of the present disclosure is implemented.

[0032] A method, device, equipment, and storage medium for inspecting and evaluating the quality of a severe convective radar identification product provided by an embodiment of the present disclosure. By obtaining a severe convective radar identification product and severe convective actual data within the coverage of a single radar station, performing time matching and spatial matching on them, and then evaluating the quality of the severe convective radar identification product according to the spatio-temporal matching result, it is a set of general quality evaluation methods applicable to all severe convective radar identification products. While the evaluation means are simple and efficient, the accuracy of the evaluation is also ensured.

[0033] It should be understood that the content described in the summary of the invention section is not intended to limit the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] With reference to the accompanying drawings and the following detailed description, the above and other features, advantages, and aspects of the embodiments of the present disclosure will become more apparent. The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. In the drawings, the same or similar reference numerals represent the same or similar elements, where:

[0035] Figure 1 Shows a flowchart of a method for inspecting and evaluating the quality of a severe convective radar identification product according to an embodiment of the present disclosure;

[0036] Figure 2 Shows a block diagram of a device for inspecting and evaluating the quality of a severe convective radar identification product according to an embodiment of the present disclosure;

[0037] Figure 3 Shows a schematic block diagram of an exemplary electronic device capable of implementing the embodiments of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present disclosure. Apparently, the described embodiments are some, but not all, of the embodiments of the present disclosure. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present disclosure without creative efforts shall fall within the scope of protection of the present disclosure.

[0039] In addition, the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after.

[0040] Figure 1 The flowchart of a method 100 for quality inspection and evaluation of severe convective radar identification products according to an embodiment of the present disclosure is shown. The method 100 includes:

[0041] Step 110, obtaining severe convective radar identification products.

[0042] In some embodiments, the severe convective radar identification products are user alarm information products. For example, various alarm products for severe convective weather of single- and dual-polarization weather radars, including hail, thunderstorm with strong wind, short-term heavy precipitation, and tornado user alarm information (UAM) products. That is, the data of user alarm information products for hail, thunderstorm with strong wind, short-term heavy precipitation, and tornado weather processes.

[0043] In some embodiments, the hail alarm product is a single-station radar hail alarm product; the thunderstorm with strong wind alarm product is a single-station radar thunderstorm with strong wind alarm product; the heavy precipitation alarm product is a single-station radar short-term heavy precipitation alarm product, and so on.

[0044] Step 120, obtaining severe convective actual situation data within the coverage of a single radar station.

[0045] In some embodiments, the severe convective actual situation data is the location where severe convection occurs, the start time and end time, and the intensity level.

[0046] In some embodiments, the hail actual data includes: hail record information in the daily observation data of ground stations; the start time, end time, location, hail size, and the area where hail occurs described in the actual situation reporting system for disasters; hail occurrence information reported by online media, etc. Before inspection, the actual data needs to be preprocessed. If the time and location of hail occurrence described in the disaster situation are consistent with the evolution of radar echoes, it is directly adopted; if the location of hail occurrence described in the disaster situation is relatively accurate but there is an error in the reported occurrence time, the reported occurrence time should be revised according to products such as radar echo intensity and VIL change. After the hail occurrence time is determined, a "time window" is established, that is, the reported start time is pushed forward by 3 volume scans (18 minutes), and the reported end time (if any) is extended backward by 2 volume scans (12 minutes). The forward push is to account for the time from hail generation to reaching the ground, and the backward extension is to account for the fact that hail lasts for a certain period and there are still certain errors in the reporting time. For example, the time and location of hail occurrence from sources such as meteorological stations, disaster situation direct reports, the Internet, and the media; the time and location of hail and tornado occurrence identified from videos taken by witnesses or on-site investigations; an echo area with a reflectivity above 65 dBZ (especially extending to heights above the 0°C level) on the radar echo in sparsely populated areas, and the radar observes triple-bounce scattering, sidelobe echoes, or a "V" notch (C-band), which can be used as one of the indication characteristics of large hail, etc.

[0047] In some embodiments, the severe thunderstorm wind actual data: the hourly or minute-level wind data of multiple stations selected from the ground observation business assessment stations in the site data actual situation. The wind data of the stations needs to be data after quality control or correction. Select the wind element and obtain the time when the strong wind appears. Note: The wind data must be typical severe thunderstorm wind data, excluding the wind data caused by cold air and typhoons, except for the wind caused by the squall line in front of the stage. Before inspection, the actual data needs to be preprocessed. Select to perform spatio-temporal matching between the ground strong wind stations and the radar reflectivity factor, and screen the strong wind stations within the range of ≥35 dBZ as the effective test true values, and read the time when the strong wind appears and match it with the severe thunderstorm wind identification time. For example, the strong wind data of the ground automatic meteorological observation station with an hourly maximum wind ≥17.2 m / s and a 2-minute average wind ≥16 m / s after quality control, excluding non-severe thunderstorm wind data such as cold air wind, terrain (such as high mountains) wind, and gust front wind separated from the parent body; the radar radial velocity data after quality control, check the strong wind core (radial velocity ≥20 m / s) and divergent velocity pairs (spacing ≤4 km, velocity difference >10 m / s) corresponding to a strong echo area (≥40 dBZ) at low elevation angles or heights below 1 km, which can be used as the actual situation of severe thunderstorm wind without ground measurement stations, and it is recorded that there is a strong wind above level 8 on the ground.

[0048] In some embodiments, for the actual heavy precipitation data: the actual site data selects the hourly precipitation or minute-level precipitation data of multiple stations of the ground observation business assessment stations, and the precipitation site data needs to be data after quality control or correction.

[0049] In some embodiments, for the actual tornado data: the actual observations or disaster investigation data observed by meteorological observation stations, etc.

[0050] Step 130, perform time matching and space matching on the severe convective radar identification product and the severe convective actual data.

[0051] In some embodiments, performing time matching and space matching on the severe convective radar identification product and the severe convective actual data includes: determining a time window according to the start time and / or end time; taking the location where severe convection occurs as the center and a preset length as the radius to determine the meteorological actual area within the time window as the actual occurrence area; determining whether there is an alarm message sent by the severe convective radar identification product (for example, the user alarm information product) within the actual occurrence area. If so, the matching is successful.

[0052] 1) Hail:

[0053] Time matching: At a certain location, taking the volume scan closest to the start time of hail as the reference, push forward 3 volume scan times as the start of the "true value" time window for hail warning, and push backward 2 volume scan times as the end of the time window. If there is an actual end time of hail, use the end time as the end of the warning "true value" time window. Space matching: After establishing the "true value" time window for hail warning, adopt the point-to-face neighborhood test method. With the actual position of hail as the center, search for user warning information hail warning areas within a radius of 20 km, and conduct space matching between the actual hail and the warning area. When there is user warning information hail warning within a radius of 20 km centered on the actual position of hail, it is rated as a hit; when there is no hail warning within a radius of 20 km centered on the actual position of hail, it is rated as a miss; for the remaining hail warnings, that is, there is no actual hail in the hail warning area, it is rated as a false alarm. For example, within 20 km, when there is both hail warning and actual hail, it is rated as "hit", recorded as 1 hit, and the number of hits is recorded as such (but when calculating the hit rate, if there is more than 1 hit, the hit rate is 100%); for a certain time period, if there are two or more warning points within a radius of 20 km, count 1 warning point as 1 hit, and the remaining warning points are all rated as false alarms, and the number of false alarm points is rated as such; if there are no warning points within a radius of 20 km, it is rated as a miss, and the number of actual points missed is rated as such. Attribution of hit and false alarm distance bins: Identify which distance bin the point is located in and classify it into that distance bin. Attribution of miss distance bins is based on the distance bin where the actual position is located. Note: When calculating the hit rate and miss rate, as long as there is 1 hit within the time window and within a radius of 20 km, then the hit rate is 100% and the miss rate is 0%.

[0054] 2) Thunderstorm gales:

[0055] If the actual data uses hourly maximum wind data:

[0056] Time matching: Search for the time when the ground hourly maximum wind of 17.2 m / s appears 1 hour after the whole hour moment of the wind data. Arrange all the times as the time period when the wind appears, and then push forward 2 volume scan times and backward 1 volume scan time from this time period as the true value of the evaluation time window.

[0057] Space matching: With the identified recognition point as the center, search within a radius of 20 km to see if there is a high wind station within this range.

[0058] Matching method:

[0059] Centered on the recognition point, within a range of 20 km, if there is 1 or more strong wind stations on the ground within this range, then this recognition point is recorded as being hit once. The wind at the ground truth station belongs to the corresponding level. If there are two or more levels, it is classified into the highest level. The recognition point belongs to the corresponding distance range. Severe thunderstorm winds are evaluated by wind speed level, divided into levels 8 - 9, 10 - 11, and ≥12 for evaluation.

[0060] Except for the above-mentioned strong wind stations that are hit, for the remaining strong wind stations, centered on this station, within a search radius of 20 km, if not recognized, it is recorded as a missed report once, and the number of stations is recorded as the number of missed reports (station spacing is greater than 20 km). The wind at the ground truth station belongs to the corresponding level, and the ground truth station belongs to the corresponding distance range.

[0061] Centered on the recognition point, within a range of 20 km, if there is no strong wind station within this range, then this recognition point is recorded as a false alarm once; the recognition point belongs to the corresponding distance range. The number of false alarms is counted regardless of intensity level, but by distance level. The number of false alarms + the number of hits = all recognition points.

[0062] If the actual data uses minute-level strong wind data (1 minute):

[0063] If the ground truth data uses minute-level data, in terms of time matching, the time of the recognition product should be consistent with the time of the ground minute-level strong wind stations. The strong wind station is defined as 16.5 m / s. For example: if the time of the recognition product is 3:00 and the next volume scan time is 3:06, then the maximum value of each strong wind station between 3:00 - 3:06 is selected as the actual data at 3:00 for the ground truth and matched with the data of the recognition product at 3:00. The spatial matching is the same as above.

[0064] 3) Short-time heavy precipitation:

[0065] The point-to-area neighborhood test method is adopted.

[0066] If the actual data uses hourly precipitation data at the whole hour:

[0067] Time matching: The recognition time of short-time heavy precipitation should be consistent with the time of the hourly precipitation data of the ground precipitation truth. For example: if the hourly precipitation data of the ground precipitation truth is at 03:00, the recognition time of short-time heavy precipitation is 03:00 or the volume scan time closest to 03:00.

[0068] Spatial matching: Based on the recognized area, search whether there are heavy precipitation stations within this recognized area for matching.

[0069] Matching method:

[0070] Within the identified area, if there is one or more rainfall stations with a rainfall of ≥20 mm / h on the ground within this area, then this identified area is recorded as being hit once. The rainfall level of the ground truth station belongs to which level is classified into that level. If there are more than two levels, it is classified into the largest level. The distance range to which the center point within the identified area belongs is classified into that distance range. At the same time, for the matched pairs that are hit, calculate the error value (add an error index for short-duration heavy rainfall), that is, the rainfall value at the identified point minus the value of the largest rainfall station within this identified area.

[0071] Excluding the above-mentioned rainfall stations that are hit (stations within the identified area), for the remaining rainfall stations, with this station as the center, within a search radius of 20 km, if not identified, it is recorded as a missed detection once, and the number of stations is recorded as the number of missed detections (station spacing is 20 km). The rainfall level of the ground truth station belongs to which level is classified into that level, and the distance range to which the ground truth station is located is classified into that distance range.

[0072] With the identified point as the center, within 20 km, if there is no rainfall station with a rainfall of more than 20 mm / h within this area, then this identified point is recorded as a false alarm once; the number of false alarms + the number of hits = all identified points.

[0073] If the ground truth data uses minute-level precipitation data:

[0074] Time matching: The time of the short-duration heavy rainfall identification product should be consistent with the time of the minute-level precipitation data of the ground precipitation truth.

[0075] For example: The time of the short-duration heavy rainfall identification product is 03:18, and the time of the ground precipitation truth is 03:18 + 1 volume scan time, that is, 03:24, and the rainfall value is the cumulative rainfall within the past 1 hour at 03:24.

[0076] Among them, the distance classification for short-duration heavy rainfall: For the S-band, it is 0 - 50 km (<50 km), 50 - 100 km, 100 - 230 km (>100 km); for the C-band, it is 0 - 50 km (<50 km), 50 - 100 km, 100 - 150 km (>100 km). The classification for short-duration heavy rainfall is evaluated according to 20 - 49.9 mm / h, 50 - 69.9 mm / h, ≥70 mm / h.

[0077] 4) Tornado:

[0078] Time matching: Determine the "time window" when a tornado occurs based on the start time and end time of the tornado described in the real-time data or disaster investigation data. The determination method is as follows: at a certain location, taking the volume scan closest to the start time of the tornado as the reference, push forward 2 volume scan times as the start of the "true value" time window for tornado warning, and push backward 1 volume scan time as the end of the time window. If there is a real-time end time of the tornado, use the end time as the end of the "true value" time window for warning.

[0079] Spatial matching: After establishing the "true value" time window for tornado warning, adopt the point-to-face neighborhood test method. With the actual location of the tornado as the center, search for user warning information within a radius of 10 km for the tornado warning area, and conduct spatial matching between the actual tornado situation and the warning area. It is also possible to search for the user warning information within the range described in the real-time data or disaster investigation data for the actual occurrence range of the tornado for the tornado warning area, and conduct spatial matching between the actual tornado situation and the warning area. The tornado assessment distances are divided into two levels: 10 - 59 km and 60 - 100 km.

[0080] When there is user warning information for a tornado within a radius of 10 km centered on the actual location of the tornado, or when searching within the range described in the real-time data or disaster investigation data for the actual occurrence range of the tornado and there is user warning information for a tornado, it is rated as a hit; when there is no warning information, it is rated as a miss; if there is no actual tornado but there is user warning information for a tornado, it is rated as a false alarm.

[0081] Step 140, conduct quality assessment on the severe convective radar recognition product according to the spatio-temporal matching result.

[0082] In some embodiments, the quality of the strong convection radar identification product is evaluated according to the time-space matching result, including: determining the hit rate of the strong convection radar identification product according to the number of successful matches; and / or determining the missed alarm rate and false alarm rate of the strong convection radar identification product according to the number of failed matches; and / or determining the alarm time advance according to the earliest time of the occurrence of strong convection in real situation and the first alarm time of the strong convection radar identification product; and / or, performing a moving path deduction according to the moving direction and speed of the strong convection radar identification product that has been successfully matched, and judging whether the corresponding strong convection in real situation occurs in the moving path area; and determining the correctness of the moving path evaluation of the strong convection radar identification product according to the judgment result. Specifically, the number of hits, missed alarms and false alarms H, M, FA are counted, and finally the total hit rate (recognition rate), missed alarm rate, false alarm rate and critical success index (accuracy) are given. Evaluation of hail alarm time advance: For the case where both ground reality and radar identification have hail, at a certain location, take the body scan closest to the start time of the live hail as the benchmark, push forward 3 body scans, and push back 2 body scans as the "true value" of the hail time window. In this time window, search for the time when the live hail first starts within a radius of 20km where the ground hail occurs (it is possible that there will be two or more ground reports of hail within 20km. If there is only one ground report of hail, the start time of the ground report is used as the hail start time). Use this time to subtract the hail alarm time of the first user alarm information in the area to obtain the hail alarm time advance, and finally calculate the average advance alarm time based on the number of evaluation cases. Of course, the accuracy evaluation of the user alarm product should give the values ​​of each indicator at different distances, and the situation of different distance levels should be analyzed in detail during the analysis; the alarm time advance should give specific values; the moving path should give correct and incorrect results. Specifically, the quality assessment process of the user warning information products under four types of severe convective weather is as follows (it should be noted that the level of the individual case should be filled in):

[0083] 1) Hail: For the evaluation of individual cases, try to select hail cases that occur at different distances. Hail distance classification: S band 0-50km (<50km), 50-100km, 100-230km (>100km); C band 0-50km (<50km), 50-100km, 100-150km (>100km). After calculating the number of hits, false alarms, and missed alarms for each hail case, and finding the corresponding hit rate, false alarm rate, and critical success index (accuracy), fill in Table 1. If multiple radars are involved in the same case, they must also be listed, and then the total number of hits, false alarms, and missed alarms is accumulated and counted, and the hit rate, false alarm rate, and critical success index (accuracy) are calculated and filled in Table 2. Finally, according to the number of selected cases, the various indicators are calculated and filled in Table 3.

[0084] Table 1: Inspection and Evaluation Results of Hail Alarm Products for XX Cases (XX Radar Stations)

[0085]

[0086]

[0087] Table 2: Inspection and Evaluation Results of Hail Alarm Products (Summary)

[0088]

[0089] Table 3: Inspection and Evaluation Results of Hail Alarm Products

[0090]

[0091] 2) Thunderstorm gales: Graded evaluation is carried out according to wind speed levels. The wind speed levels are shown in Table 4, and the evaluation is carried out for levels 8 - 9, 10 - 11, and ≥12. Classification of thunderstorm gale distances: For S - band, 0 - 50 km (<50 km), 50 - 100 km, 100 - 230 km (>100 km); for C - band, 0 - 50 km (<50 km), 50 - 100 km, 100 - 150 km (>100 km).

[0092] Table 4: Table for Classifying Ground Gale Intensity Levels

[0093] Classification Wind speed (m / s) Grade 8 17.2≤V≤20.7 Grade 9 20.8≤V≤24.4 Grade 10 24.5≤V≤28.4 Grade 11 28.5≤V≤32.6 Grade 12 32.7≤V≤36.9 > Grade 12 V≥37.0

[0094] Calculate the probability of severe thunderstorm winds based on their characteristics, generate severe thunderstorm wind warning products, trigger a severe thunderstorm wind warning when the probability reaches 55%, and give the fitted warning area, its moving direction and speed. Time lead assessment: For cases where there are severe thunderstorm winds in both ground truth and radar identification, at a certain location, select the earliest time among the first occurrence time of the 2-minute average wind ≥ 16 m / s and the hour maximum wind occurrence time as the ground strong wind start time. Taking the volume scan closest to this time as the reference, push forward 3 volume scans and backward 2 volume scan times as the "true value" of the severe thunderstorm wind warning time window. Within this time window, search for the severe thunderstorm wind warning time of the first user warning information within a radius of 20 km from the ground strong wind occurrence location. Subtract the first severe thunderstorm wind warning time from the ground strong wind start time to obtain the severe thunderstorm wind warning lead time. Finally, calculate the average early warning time based on the number of evaluated cases. Severe thunderstorm wind moving path assessment: Extrapolate the severe thunderstorm wind warning area for 1 hour based on the moving direction and speed of the severe thunderstorm wind warning area. If severe thunderstorm winds occur within the moving path area, it is rated that the moving path of the severe thunderstorm wind warning is correct. The assessment of the severe thunderstorm wind moving path can adopt qualitative assessment. The accuracy assessment of severe thunderstorm winds should give the index values at different wind speed levels and different distances. When analyzing, the situations at different wind speed levels and different distance levels should be analyzed in detail; the warning time lead should give specific values; the moving path should give correct and incorrect results. Try to select severe thunderstorm wind cases at different wind speed levels and different distance levels for the selected cases. After calculating the number of hits, false alarms, and missed alarms for each severe thunderstorm wind case and obtaining the corresponding hit rate, false alarm rate, and critical success index (accuracy rate), fill them into Table 5. If there are multiple radars involved in the same case, they should also be listed, and then accumulate and count the total number of hits, false alarms, and missed alarms, calculate the hit rate, false alarm rate, and critical success index (accuracy rate), and fill them into Table 6. Finally, based on the number of selected cases, summarize and calculate each index and fill it into Table 7.

[0095] Table 5: Inspection and Evaluation Results of Severe Thunderstorm Wind Warning Products for XX Cases (XX Stations)

[0096]

[0097]

[0098] Table 6: Inspection and Evaluation Results of Severe Thunderstorm Wind Warning Products for XX Cases (Summary)

[0099]

[0100]

[0101] Table 7: Inspection and Evaluation Results of Severe Thunderstorm Wind Warning Products

[0102]

[0103] 3) Short-term heavy precipitation:

[0104] Distance classification of short-term heavy precipitation: S-band 0 - 50 km (<50 km), 50 - 100 km, 100 - 230 km (>100 km); C-band 0 - 50 km (<50 km), 50 - 100 km, 100 - 150 km (>100 km).

[0105] Rainfall grading for short-term heavy precipitation: It is graded and evaluated according to 20 - 49.9 mm / h, 50 - 69.9 mm / h, and ≥70 mm / h. For example, taking the level of 20 - 49.9 mm / h as an example, if there is a rain gauge station with a rainfall of 20 - 49.9 mm in one hour within the heavy precipitation warning area, it is rated as a hit; if there is no rain gauge station with a rainfall of 20 - 49.9 mm in one hour within the heavy precipitation warning area, it is rated as a false alarm; if the rain gauge station with a rainfall of 20 - 49.9 mm in one hour does not fall within the heavy precipitation warning area, it is rated as a missed alarm. Count the number of hits, false alarms, and missed alarms, and calculate the hit rate (recognition rate), false alarm rate, and critical success index (accuracy rate). The short-term heavy precipitation warning product is generated based on the 1-hour precipitation OHP estimated by radar (product No. 169 is used for dual-polarization radar, and product No. 78 is used for single-polarization radar). If the continuous area with an estimated 1-hour precipitation ≥20 mm exceeds 30 km², a short-term heavy precipitation warning is triggered, and the fitted elliptical warning area, its moving direction, and speed are given. At a certain location, taking the volume scan closest to the first occurrence time of ground 1-hour precipitation ≥20 mm as the reference, 3 volume scan times are pushed forward and 2 volume scan times are pushed backward as the "true value" of the short-term heavy precipitation warning time window. Within this time window, search for the first heavy precipitation warning time within a radius of 20 km from the ground heavy precipitation occurrence location. Subtract the first user warning information heavy precipitation warning time from the first heavy precipitation actual occurrence time to obtain the heavy precipitation warning lead time. Finally, calculate the average lead warning time based on the number of evaluated cases. Extrapolate the heavy precipitation warning area by 1 hour according to the moving direction and speed of the heavy precipitation warning area. If heavy precipitation occurs within the moving path area, it is rated as the correct moving path of the heavy precipitation warning. The evaluation of the heavy precipitation moving path can adopt qualitative evaluation. The accuracy evaluation of the heavy precipitation warning product should give the index values at different precipitation levels and different distance levels. When analyzing, the situations at different precipitation levels and different distance levels should be analyzed in detail; the specific value of the warning time lead should be given; the correct and incorrect results of the moving path should be given. When selecting evaluated cases, try to select heavy precipitation cases at different precipitation levels and different distance levels; select 3 heavy precipitation cases per month. If there is no heavy precipitation weather, no evaluation is conducted. However, for radars in the western region with very few cases of ≥20 mm / h, the evaluation can be carried out according to the local heavy precipitation standard. After calculating the number of hits, false alarms, and missed alarms for each heavy precipitation case and obtaining the corresponding hit rate, false alarm rate, and critical success index (accuracy rate), fill them into Table 8. If multiple radars are involved in the same case, they should also be listed, and then the total number of hits, false alarms, and missed alarms should be accumulated and counted, and the hit rate, false alarm rate, and critical success index (accuracy rate) should be calculated and filled into Table 9. Finally, according to the number of selected cases, summarize and calculate each index and fill it into Table 10.

[0106] Table 8: Test and Evaluation Results of Short-Term Heavy Precipitation Warning Products for XX Cases (XX Stations)

[0107]

[0108] Table 9: Inspection and Evaluation Results of Short-term Heavy Rainfall Alarm Products for XX Cases (Summary)

[0109]

[0110]

[0111] Table 10: Inspection and Evaluation Results of Short-term Heavy Rainfall Alarm Products

[0112]

[0113]

[0114] 4) Tornado:

[0115] Taking into account the echo characteristics of tornado storms, multiple radar characteristic quantities, and the requirements for TVS identification, the tornado evaluation distance is divided into two levels: 10 - 59 km and 60 - 100 km. When there is a tornado alarm in the user alarm information within a 10 km radius centered on the actual location of the tornado, or when searching for tornado alarms in the user alarm information within the range described by the actual situation data or disaster investigation data for the actual occurrence range of the tornado, it is rated as a hit. When there is no alarm information, it is rated as a miss; if there is no tornado in the actual situation, but there is a tornado alarm in the user alarm information, it is rated as a false alarm. After counting the number of hits, false alarms, and misses for a certain tornado case, and calculating the corresponding hit rate, false alarm rate, and critical success index (accuracy rate), fill them into Table 11. If multiple radars are involved for the same case, they should also be listed, and then the total number of hits, false alarms, and misses is accumulated and counted. After calculating the hit rate, false alarm rate, and critical success index (accuracy rate), fill them into Table 12. Finally, according to the number of tornado cases, summarize and calculate each index and fill it into Table 13.

[0116] Table 11: Inspection and Evaluation Results of Tornado Alarm Products for XX Cases at XX Stations

[0117]

[0118] Table 12: Inspection and Evaluation Results of Tornado Alarm Products for XX Cases (Summary)

[0119]

[0120] Table 13: Inspection and Evaluation Results of Tornado Alarm Products

[0121]

[0122] The following are the relevant data (Table 14 - 20) on the evaluation results of user alarm information products for four types of severe convections: hail, thunderstorm gales, short-term heavy precipitation, and tornadoes:

[0123] Table 14: Overall Evaluation Results of National Hail Alarm Products

[0124]

[0125] Table 15: Overall Evaluation Results of Hail Alarm Products in North China Region

[0126]

[0127] Table 16: Overall Evaluation Results of Hail Alarm Products in East China Region

[0128]

[0129] Table 17: Overall Evaluation Results of National Thunderstorm Gale Alarm Products

[0130]

[0131] Table 18: Overall Evaluation Results of National Short-Term Heavy Precipitation Alarm Products

[0132]

[0133]

[0134] Table 19: Overall Evaluation Results of National Tornado Alarm Products

[0135]

[0136] In the above embodiments, the calculation formulas for the hit rate, miss rate, false alarm rate, and critical success index are as follows:

[0137] Hit rate:

[0138] Miss rate:

[0139] False alarm rate:

[0140] Critical success index:

[0141] Where: H, M, and FA respectively represent the number of hits, misses, and false alarms.

[0142] Therefore, in combination with the corresponding actual situation data, the probabilities of hits, misses, and false alarms can be accurately calculated. In addition, the problem of difficult to accurately evaluate the recognition lead due to the lag in data collection of the actual situation occurrence time is also solved. It has wide applicability and is simple and efficient.

[0143] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present disclosure is not limited by the described action sequence, because according to the present disclosure, certain steps can be in other sequences or performed simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to the present disclosure.

[0144] The above is the introduction of the method embodiments. The following further illustrates the solution of the present disclosure through device embodiments.

[0145] Figure 2 The block diagram of the severe convective radar recognition product quality inspection and evaluation device 200 according to an embodiment of the present disclosure is shown. As Figure 2 shown, the device 200 includes:

[0146] A data acquisition module 210, configured to acquire severe convective radar recognition products;

[0147] The data acquisition module 210 is further configured to acquire severe convective actual situation data within the coverage of a single radar station;

[0148] A matching module 220, configured to perform time matching and space matching on the severe convective radar recognition product and the severe convective actual situation data;

[0149] An evaluation module 230, configured to perform quality evaluation on the severe convective radar recognition product according to the spatio-temporal matching result.

[0150] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the described modules can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0151] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device and a readable storage medium.

[0152] Figure 3FIG. 0 shows a schematic block diagram of an electronic device 300 that can be used to implement embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as, for example, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, for example, personal digital processors, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely exemplary and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0153] The electronic device 300 includes a computing unit 301 that can perform various appropriate actions and processes in accordance with a computer program stored in the ROM 302 or a computer program loaded from the storage unit 308 into the RAM 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 can also be stored. The computing unit 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. The I / O interface 305 is also connected to the bus 304.

[0154] A plurality of components in the electronic device 300 are connected to the I / O interface 305, including: an input unit 306, such as a keyboard, a mouse, etc.; an output unit 307, such as various types of displays, speakers, etc.; a storage unit 308, such as a magnetic disk, an optical disk, etc.; and a communication unit 309, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 309 allows the electronic device 300 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0155] The computing unit 301 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 301 executes the various methods and processes described above, such as method 100. For example, in some embodiments, method 100 can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 308. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 300 via the ROM 302 and / or the communication unit 309. When the computer program is loaded into the RAM 303 and executed by the computing unit 301, one or more steps of the method 100 described above can be executed. Alternatively, in other embodiments, the computing unit 301 can be configured to execute method 100 in any other suitable manner (e.g., by means of firmware).

[0156] Various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuitry, integrated circuit systems, field-programmable gate arrays (FPGA), application-specific integrated circuits (ASIC), application-specific standard products (ASSP), system-on-a-chip systems (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special or general-purpose programmable processor, receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0157] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program code can be executed entirely on the machine, partially on the machine, as an independent software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0158] In the context of this disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0159] In order to provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0160] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0161] A computer system can include a client and a server. The client and the server are generally far apart from each other and typically interact through a communication network. The relationship between the client and the server is generated by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, can also be a server of a distributed system, or a server incorporating a blockchain.

[0162] It should be understood that the various forms of processes shown above can be used, with steps reordered, added or deleted. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and no limitations are imposed herein.

[0163] The above specific embodiments do not constitute a limitation on the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub - combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the protection scope of this disclosure.

Claims

1. A method for inspecting and evaluating the quality of severe convective radar identification products, characterized in that, including: Obtaining severe convective radar recognition products; Obtaining the actual severe convective data within the coverage of a single radar station; Performing time matching and spatial matching on the severe convective radar recognition products and the actual severe convective data; Evaluating the quality of the severe convective radar recognition products according to the spatio-temporal matching results.

2. The method according to claim 1, wherein the severe convective radar recognition products are user alarm information products; the actual severe convective data are the location where severe convection occurs, the start time and end time, and the intensity level; Performing time matching and spatial matching on the user alarm information products and the actual severe convective data includes: Determining a time window according to the start time and / or end time; Taking the location where severe convection occurs as the center and a preset length as the radius, determining the meteorological actual situation area within the time window as the actual occurrence area; Determining whether there is an alarm message sent by the user alarm information product within the actual occurrence area. If so, the matching is successful.

3. The method according to claim 2, wherein Evaluating the quality of the user alarm information products according to the spatio-temporal matching results includes: Determining the hit rate of the user alarm information products according to the number of successful matches; and / or, determining the false alarm rate and missed alarm rate of the user alarm information products according to the number of failed matches; and / or, determining the alarm time lead according to the earliest time when severe convection actually occurs and the first alarm time of the user alarm information product; and / or, performing a movement path deduction according to the movement direction and speed of the storm cell given by the user alarm information product, and determining whether there is a corresponding actual severe convective situation within the movement path area; determining the correctness of the movement path evaluation of the user alarm information product according to the judgment result.

4. The method according to claim 2, wherein the user alarm information products are divided into four categories: hail, thunderstorm with strong wind, short-term heavy precipitation, and tornado.

5. The method according to claim 1, wherein The method further includes: Preprocessing the actual severe convective data.

6. An apparatus for inspecting and evaluating the quality of a severe convective radar identification product, characterized in that, including: A data acquisition module for obtaining severe convective radar recognition products; The data acquisition module is also used to obtain the actual severe convective data within the coverage of a single radar station; A matching module for performing time matching and spatial matching on the severe convective radar recognition products and the actual severe convective data; An evaluation module for evaluating the quality of the severe convective radar recognition products according to the spatio-temporal matching results.

7. An electronic device, characterized in that, including: At least one processor; and A memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1-5.

8. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to execute the method according to any one of claims 1-5.

Images